See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/292318981

Teaching quantum physics at introductory level: a sum over paths approach

Thesis · January 2016

CITATION READS

1 795

1 author:

Massimiliano Malgieri
University of Pavia
56 PUBLICATIONS   259 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Teaching introductory quantum physics through Feynman's sum over paths approach View project

Low cost spectroscopy View project

All content following this page was uploaded by Massimiliano Malgieri on 13 July 2016.

The user has requested enhancement of the downloaded file.

 Teaching quantum physics
at introductory level:
a sum over paths approach

Massimiliano Malgieri

Submitted to the Graduate School in Physics

in partial fulfillment of the requirements
for the degree of
Doctor of Philosophy in Physics
at the
University of Pavia

Advisor: Prof. Anna De Ambrosis

To Paola, Riccardo and Caterina.
Contents

Table of contents vi

List of Figures vii

1 Introduction 1
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Main research questions . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Dissertation outline . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Reference guidelines from physics education research 7

2.1 The Pavia approach to physics education . . . . . . . . . . . . . 7
2.1.1 The design of teaching-learning sequences . . . . . . . . 7
2.1.2 Innovation in teaching modern physics: the case of rela-
tivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.3 The use of ICT in education . . . . . . . . . . . . . . . . 10
2.1.4 The role of history of physics . . . . . . . . . . . . . . . 10
2.2 The concept of appropriation . . . . . . . . . . . . . . . . . . . 11
2.3 Conceptual change in the perspective of framework theories . . . 13
2.4 Quantitative data analysis using knowledge integration rubrics . 15

3 A panorama of research on teaching quantum physics 19

3.1 Research on student difficulties . . . . . . . . . . . . . . . . . . 20
3.1.1 Conceptual difficulties and students’ models . . . . . . . 20
3.1.2 Difficulties with ideas from classical physics and mathe-
matical concepts . . . . . . . . . . . . . . . . . . . . . . 30
3.2 Overview of educational perspectives . . . . . . . . . . . . . . . 33
3.2.1 General approaches to the disciplinary content . . . . . . 33
3.2.2 Epistemological perspectives . . . . . . . . . . . . . . . . 37
3.2.3 Complementary strategies and tools . . . . . . . . . . . . 40

4 The sum over paths approach 43

4.1 Fundamentals of Feynman’s path integral formulation . . . . . . 43
4.1.1 Classical and quantum probabilities . . . . . . . . . . . . 43

iii
 CONTENTS

4.1.2 Intuitive approach to the path integral formulation . . . 44

4.1.3 Postulates of the path integral formulation . . . . . . . . 45
4.1.4 From the propagator to the wave function . . . . . . . . 47
4.1.5 The correspondence principle . . . . . . . . . . . . . . . 48
4.1.6 Equivalence to the Schrödinger equation . . . . . . . . . 48
4.2 Sum over paths for photons: Feynman’s QED . . . . . . . . . . 50
4.2.1 QED chapters 1 and 2: sum over paths for the monochro-
matic photon . . . . . . . . . . . . . . . . . . . . . . . . 51
4.2.2 QED chapters 3 and 4: Scalar electrodynamics and the
Klein-Gordon equation . . . . . . . . . . . . . . . . . . . 54
4.3 Educational proposals based on the sum over paths approach . . 56
4.3.1 The Taylor et al. approach . . . . . . . . . . . . . . . . . 56
4.3.2 Ogborn, Dobson and the Advancing Physics AS project . 59
4.3.3 Rinaudo and the the proposal of Torino . . . . . . . . . 61
4.3.4 Other proposals . . . . . . . . . . . . . . . . . . . . . . . 62
4.4 Treatment of time independent problems . . . . . . . . . . . . . 63
4.5 Sum over paths methods for the Green function . . . . . . . . . 64
4.5.1 The Van Vleck - Gutzwiller propagator . . . . . . . . . . 66
4.5.2 Sum over paths representation of the Green function for
piecewise constant potentials . . . . . . . . . . . . . . . . 66
4.6 One dimensional bound systems in piecewise constant potentials 67
4.6.1 Infinite square well . . . . . . . . . . . . . . . . . . . . . 67
4.6.2 Finite square well . . . . . . . . . . . . . . . . . . . . . . 69
4.6.3 Asymmetric infinite square well . . . . . . . . . . . . . . 72
4.7 One dimensional open system: square barrier tunneling . . . . . 74
4.8 One dimensional smooth potential: the harmonic oscillator . . . 75
4.8.1 Comments on the results and their educational translation 78
4.9 Sum over paths as an interpretative language for modern exper-
iments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

5 The design of our proposal 81

5.1 Main educational choices . . . . . . . . . . . . . . . . . . . . . . 82
5.1.1 Making the conceptual and epistemological structure of
the theory transparent. . . . . . . . . . . . . . . . . . . . 82
5.1.2 Discussing modern experiments . . . . . . . . . . . . . . 83
5.1.3 Addressing mathematical difficulties of students . . . . . 85
5.1.4 Visualization and the use of simulations . . . . . . . . . 86
5.1.5 Pursuing appropriation . . . . . . . . . . . . . . . . . . . 88
5.1.6 Including a practical experimental dimension . . . . . . . 88
5.2 The 2014 version . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.2.1 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.2.2 The sum over paths method in wave optics . . . . . . . . 90
5.2.3 The photon concept . . . . . . . . . . . . . . . . . . . . 91
5.2.4 The Feynman photon model . . . . . . . . . . . . . . . . 92

iv
CONTENTS

5.2.5 Using the Feynman approach for treating conceptual and

foundational issues . . . . . . . . . . . . . . . . . . . . . 93
5.2.6 Massive particles . . . . . . . . . . . . . . . . . . . . . . 97
5.2.7 The limit of geometrical optics and classical mechanics:
the correspondence principle . . . . . . . . . . . . . . . . 98
5.2.8 Quantization in the sum over paths approach . . . . . . 98
5.2.9 Comments on the structure of the 2014 sequence . . . . 99
5.3 The 2015 sequence . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.3.1 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.3.2 Initial experiments . . . . . . . . . . . . . . . . . . . . . 101
5.3.3 The photon concept . . . . . . . . . . . . . . . . . . . . 102
5.3.4 Feynman’s photon model . . . . . . . . . . . . . . . . . . 103
5.3.5 Conceptual and foundational aspects, first part . . . . . 103
5.3.6 Massive particles . . . . . . . . . . . . . . . . . . . . . . 104
5.3.7 Conceptual and foundational aspects, second part . . . . 104
5.3.8 Open systems and the tunnel effect . . . . . . . . . . . . 107
5.3.9 Bound systems and quantization rules . . . . . . . . . . 109
5.3.10 Measurement of the Rydberg constant . . . . . . . . . . 110
5.3.11 Comments on the 2015 sequence . . . . . . . . . . . . . . 111
5.4 Details of GeoGebra simulations . . . . . . . . . . . . . . . . . . 112
5.4.1 Interference and diffraction . . . . . . . . . . . . . . . . . 112
5.4.2 Double slit interference . . . . . . . . . . . . . . . . . . . 114
5.4.3 Refraction at an interface . . . . . . . . . . . . . . . . . 114
5.4.4 Lloyd mirror and phase loss upon reflection . . . . . . . 116
5.4.5 Mach-Zehnder interferometer . . . . . . . . . . . . . . . 118
5.4.6 The limits of geometrical optics and classical mechanics . 120
5.4.7 Simulations of time independent bound systems . . . . . 120
5.4.8 Simulations of time independent open systems . . . . . . 122
5.5 Rethinking the sequence for the 2015 test in high school . . . . 124
5.5.1 Longitudinality . . . . . . . . . . . . . . . . . . . . . . . 125
5.5.2 Multi-dimensionality . . . . . . . . . . . . . . . . . . . . 126
5.5.3 Multi-perspectiveness . . . . . . . . . . . . . . . . . . . . 127

6 Implementation and results 131

6.1 2014 test with student teachers . . . . . . . . . . . . . . . . . . 131
6.1.1 Context and organization of the study . . . . . . . . . . 131
6.1.2 Sources of data and method of analysis . . . . . . . . . . 132
6.1.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
6.2 2015 test with student teachers . . . . . . . . . . . . . . . . . . 137
6.2.1 Context and organization of the study . . . . . . . . . . 137
6.2.2 Sources of data and method of analysis . . . . . . . . . . 138
6.2.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
6.3 2015 test with high school students . . . . . . . . . . . . . . . . 149
6.3.1 Context and organization of the study . . . . . . . . . . 149
6.3.2 Sources of data and method of analysis . . . . . . . . . . 151

v
 CONTENTS

6.3.3 General development of the activities and students’ re-

sponse . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
6.3.4 Results on student’s understanding and learning difficulties156
6.3.5 Analysis of cases of appropriation . . . . . . . . . . . . . 164

7 Conclusions and future perspectives 179

7.1 Re-evaluation of research questions . . . . . . . . . . . . . . . . 179
7.2 Future expansions of this work . . . . . . . . . . . . . . . . . . . 186

A Test sheets and questionnaires 189

A.1 2014 experimentation with student teachers . . . . . . . . . . . 189
A.1.1 Pre-test . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
A.1.2 Post-test . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
A.1.3 Open question in the final exam . . . . . . . . . . . . . . 192
A.2 2015 test with student teachers . . . . . . . . . . . . . . . . . . 192
A.2.1 Initial mini-questionnaire . . . . . . . . . . . . . . . . . . 192
A.2.2 Final exam . . . . . . . . . . . . . . . . . . . . . . . . . 192
A.3 2015 test with high school students . . . . . . . . . . . . . . . . 194
A.3.1 Initial questionnaire . . . . . . . . . . . . . . . . . . . . 194
A.3.2 Guidelines for students’ argumentative paper . . . . . . . 199
A.3.3 Final test for school grading . . . . . . . . . . . . . . . . 203
A.3.4 Final test-questionnaire . . . . . . . . . . . . . . . . . . 204

B Exercises 209
B.1 Exercises for the 2014 version . . . . . . . . . . . . . . . . . . . 209
B.2 Exercises added in 2015 . . . . . . . . . . . . . . . . . . . . . . 214

C Protocol for the final interviews in the high school course 219
C.1 Final interview for the quantum physics sequence . . . . . . . . 219
C.1.1 Introduction to the interview . . . . . . . . . . . . . . . 219
C.1.2 List of questions to ask students: . . . . . . . . . . . . . 220

D Realization and calibration of a low cost spectrophotometer 223

D.1 Wavelength calibration . . . . . . . . . . . . . . . . . . . . . . . 223
D.2 Intensity calibration . . . . . . . . . . . . . . . . . . . . . . . . 225
D.3 Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227

Bibliography 229

List of Publications 263

Refereed Journal Papers . . . . . . . . . . . . . . . . . . . . . . . . . 263
Refereed Conference Papers . . . . . . . . . . . . . . . . . . . . . . . 264

vi
 CONTENTS

viii
List of Figures

3.1 Students’ inaccurate representations of the tunneling wave func-

tion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

4.1 Multiple interference setup with successive slits. . . . . . . . . . 45

4.2 The Cornu spiral. . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.3 Reproduction of Fig. 56 on p. 90 of QED. . . . . . . . . . . . . 55
4.4 Principal paths for each family in the infinite square potential
well. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.5 Principal paths for each family in the inside-outside transition
in the finite square well potential. . . . . . . . . . . . . . . . . . 71
4.6 Principal paths for the asymmetric infinite square well potential. 72
4.7 Computed densities for n = 3. . . . . . . . . . . . . . . . . . . . 78

5.1 Schematic representation of the teaching sequence. . . . . . . . . 90

5.2 Examples of animations and simple experiments for the initial
activities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.3 Schematic representation of the Zhou-Wang-Mandel experiment. 96
5.4 Experimental results for the ZWM experiment in the indistin-
guishable and distinguishable cases (reproduced from Ref. [354]) 96
5.5 Schematic representation of the teaching sequence for 2015. . . . 101
5.6 Schematic representation and Feynman diagrams for the Hong-
Ou-Mandel experiment. . . . . . . . . . . . . . . . . . . . . . . 105
5.7 Captured spectra and intensity vs. wavelength graphs for the
activity on the measurement of Rydberg’s constant. . . . . . . . 110
5.8 Sum over paths simulation of Single slit diffraction. . . . . . . . 113
5.9 Sum over paths simulation of a diffraction grating. . . . . . . . . 113
5.10 Sum over paths simulations of two slit interference. . . . . . . . 115
5.11 Sum over paths simulations of the refraction of a photon at the
interface between two materials . . . . . . . . . . . . . . . . . . 116
5.12 Simulation of the Lloyd mirror. . . . . . . . . . . . . . . . . . . 117
5.13 Simulation of the Mach-Zehnder interferometer. . . . . . . . . . 119
5.14 Simulation of a parabolic mirror. . . . . . . . . . . . . . . . . . 121

ix
 LIST OF FIGURES

5.15 Simulation of two slits interference with massive particles. . . . 121

5.16 Sum over paths simulation for the infinite square well. . . . . . . 123
5.17 Sum over paths simulation of resonant (double delta) and square
barrier tunneling. . . . . . . . . . . . . . . . . . . . . . . . . . . 124

6.1 Post test item requiring ST to compute detection probabilities. . 133

6.2 Post-test open question on single photon interpretation. . . . . . 134
6.3 Post-test multiple choice question on the role of measurement. . 135
6.4 Multiple choice question of the final test concerning the Mach-
Zehnder interferometer. . . . . . . . . . . . . . . . . . . . . . . . 144
6.5 Multiple choice question concerning the energy of the emitted
photon in an atomic transition . . . . . . . . . . . . . . . . . . . 145
6.6 Multiple choice question concerning the Mach-Zehnder interfer-
ometer with interaction-free which way measurement. . . . . . . 146

A.1 Depiction of the uncertainty principle (reproduced from Ref.

[421]). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
A.2 The standard Mach-Zehnder experiment and its results. . . . . . 205
A.3 Results for the Mach-Zehnder experiment with an intermediate
which way detector. . . . . . . . . . . . . . . . . . . . . . . . . . 206
A.4 The Merli-Missiroli-Pozzi experiment. . . . . . . . . . . . . . . . 206

B.1 Setup for the first part of the exercise. . . . . . . . . . . . . . . 210

B.2 Setup for question 1:c. . . . . . . . . . . . . . . . . . . . . . . . 210
B.3 Setup for questions 1:e, 1:f and 1:h. . . . . . . . . . . . . . . . . 211
B.4 Setup for question 1:g. . . . . . . . . . . . . . . . . . . . . . . . 211
B.5 Setup for question 1:j. . . . . . . . . . . . . . . . . . . . . . . . 212
B.6 Mach-Zehnder interferometer setup. . . . . . . . . . . . . . . . . 212
B.7 Snapshot of the Mach-Zehnder simulation. . . . . . . . . . . . . 213
B.8 Schematic representation of a two slits experiment with discrete
detectors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
B.9 Three slits setup. . . . . . . . . . . . . . . . . . . . . . . . . . . 215
B.10 Setup with multiple slits for questions 3:a, 3:b and 3:c. . . . . . 216
B.11 Setup for exercise 5. . . . . . . . . . . . . . . . . . . . . . . . . 217

D.1 Set up of the home made spectrometer and recorded spectra. . . 224
D.2 Spectra recorded using the home made spectrophotometer and
sensitivity function. . . . . . . . . . . . . . . . . . . . . . . . . . 226

x
Chapter 1
Introduction

1.1 Motivation
This dissertation deals with the general educational problem of teaching rudi-
ments of quantum physics to high school students, high school teachers, and
non specialist undergraduates. A primary motivation for this work is that the
usual formulation of the subject which is employed in undergraduate physics
courses crucially relies on the mathematical structure of Hilbert spaces, which
is not remotely accessible to students in the final year of high school. As a
consequence, the introduction of quantum mechanics at this level usually does
not go very far beyond a schematic overview of the first steps of its historical
development, i.e. of what is usually called the “old quantum theory”.
More than a century after its beginnings, quantum theory is certainly no longer
in its infancy, and has deeply shaped the scientists’ perspective on nature, well
beyond the domain in which it was born, i.e. atomic theory. As such, the cen-
tral elements of its conceptual structure certainly deserve to be introduced to
all students, not only those specializing in physics and chemistry; and in sev-
eral countries, including Italy, high school curriculum reforms are progressively
approaching this objective. There are, however, a number of risks involved in
assigning the task of teaching the basics of quantum physics to teachers who
have not been adequately prepared and formed to this aim. These risks in-
clude over-simplification and fragmentation of the content [1], and relevant
misunderstandings of its epistemological significance. In this sense, a primary
direction of our work has been (and is) in the formation of teachers.
In our work we chose to adopt Feynman’s sum over paths approach to quan-
tum physics [2, 3, 4]. Feynman’s formulation has a consolidated tradition in
education [5, 6, 7, 8, 9, 10, 11, 12] and its primary merits are its conceptual
clarity, and mathematical simplicity. Also, the sum over paths approach offers
to students the possibility of visualizing the mathematical model in simple
cases, and permits a seamless connection to, and continual comparison with,
the results and predictions of classical physics. However, the approach also has

1
 1. Introduction

weaknesses [5, 13, 14], and critical objections have been raised in the literature
against its adoption in education [15]. Building an educational reconstruction
of quantum physics using the sum over paths perspective which could bring
improvements with respect to the known limitations of the approach, and re-
spond to the criticisms that have been raised against it has been one of the
primary motivations for starting the research work which eventually led to this
dissertation.
The result of our work at this stage is not a full course in quantum physics,
not even an introductory one. As will be discussed in the conclusions (Sec-
tion 7.2), several important topics which we would like to introduce still are
not covered; and many others probably will never be, given the limited time
available both in the high school setting, and in teacher training courses. It
is instead, in the consolidated educational tradition of teaching-learning se-
quences [16, 17], a self contained learning path, attempting to provide a solid
introduction to the conceptual structure of the theory in its “state of the art”
understanding, and focusing more on those themes and threads which have
proved most fruitful for its recent developments, both in the experimental and
foundational dimensions.

1.2 Main research questions

In this section, we condense the principal directions of our work in the form
of research questions, which we have tried to address. Many of the objectives
herein enumerated will be analyzed more in depth in Section 5.1, where we will
account for the main educational choices that we took in the initial develop-
ment of our sequence, and Sections 5.3 and 5.5, where we discuss the evolution
of our sequence between the first experimentation and the successive ones.
Here, however, we present the research questions, which will be re-evaluated
in the conclusions (Chapter 7) in a non-chronological order, without distin-
guishing between those which we had present from the start, and those which
were raised as a consequence of educational experience.
One of the main objectives of our work has been to expand the scope and
possibilities of the educational use of Feynman’s approach. This has included
three principal directions.

ˆ On a conceptual level, we had as a primary reference the use that is made of

Feynman’s formulation in quantum optics [18], where it is adopted not only
as an algorithm for performing computations, but as a language to interpret
experimental setups and conceptual issues, particularly those concerning
interference not of different paths of an individual photon, but of alternative
indistinguishable processes, or histories, involving the whole experimental
setup.

ˆ On a formal level, we tried to progress in overcoming the known difficulties

of the sum over paths approach in providing an elementary treatment of

2
1.2. Main research questions

time independent quantum systems [5, 14]. To this aim, we proceeded on

the work of formulating in a compact way and translating into an elemen-
tary language the sum over paths representation of the energy-dependent
propagator (or Green function) which had been started already by Onorato
of the Pavia group before the beginning of this work [19].

ˆ At the level of didactic practice, we worked on improving over the existing

proposals using the sum over paths approach, with the aim of providing
students not only with instruments for qualitative, conceptual understand-
ing, but also with a minimal amount of formalism, allowing them to obtain
quantitative answers in relevant physical problems. In this sense, an impor-
tant part of our work was on the design of exercises and problems specif-
ically aimed at the understanding of the sum over paths formalism; and
on establishing, starting from Feynman’s prospective, universal concepts of
quantum theory, such as wave functions, stationary states and energy lev-
els, in such a way to permit to teachers and students to deal with general
purpose exercises, such as those reported in textbooks.
The three previous objectives can be translated in the following research ques-
tions:
Q1: Can we transfer the way in which the Feynman approach is used by scien-
tists, particularly in the field of quantum optics, to the high school context,
making it become a language to discuss experimental setups and concep-
tual issues, first and foremost those concerning wave particle duality?

Q2: Can we, also through an in-depth investigation on the disciplinary content,
perform a suitable educational translation of the sum over paths represen-
tation of the Green function, obtaining a consistent method for dealing
with time independent problems from a sum over paths perspective?

Q3: Is the sum over paths formulation at least as efficient as other approaches
in enabling students to find quantitative answers for typical introductory
problems and exercises in quantum physics?
Multiple studies [20, 21, 22, 23, 24] have reported that, in their first approach
to quantum physics, students often produce hybrid, inconsistent quantum-
classical models. This phenomenon has been connected by some authors
[23, 24, 25] to the excessive importance given in instruction to semiclassical
models of the “old quantum theory” in the context of an historical approach.
In our work, we have interpreted this educational problem in terms of Vosni-
adou’s model for conceptual change [26, 27], involving competition and con-
tinual interplay between two different “framework theories”, which can lead to
inconsistencies, synthetic conceptions, and other forms of hybridization. From
such theoretical perspective we have drawn several indications [28] on the de-
velopment of our teaching-learning sequence, which can be expressed by the
following research questions:

3
 1. Introduction

Q4: Can we, by properly organizing the sequential presentation of the content
in a non-historical order which takes into account the possible generation
of inconsistencies in students’ models, prevent the formation of hybrid,
synthetic conceptions?

Q5: Is the introduction, at precisely selected points in the sequence, of modern

experiments, which sharply highlight the non classical features of quantum
theory, an effective strategy for contributing to achieve the above goal?
The possibility to provide a visualization of the quantum model has been enu-
merated among the strong points of the sum over paths approach; and we, as
other authors have done, have tried to exploit this characteristic to its fullest
extent, by producing a large number of simulations of relevant problems. The
effectiveness of interactive simulations in sustaining student’s learning of physi-
cal concepts is widely acknowledged. However, there is no consensus, nor there
has ever been, among scientists on the role of visualization in the understand-
ing of quantum physics. This debate started already at the first development
stages of the theory, with the controversy between Schrödinger and Heisenberg
[29, 30, 31] on the respective anschaulichkeit1 of the wave and matrix formula-
tions. In the physics education community, some authors have argued against
providing to students any form of visualization of quantum objects [32, 33].
More specifically, the kind of visualization which is provided by the sum over
paths approach has been criticized, for being too closely related to the classical
concept of trajectory [15]. Other authors, as we will more extensively discuss
in the review on this subject in Sec. 3.2.3.2 have held different views, but
certainly the problem of providing forms of visualization which do not mislead
students into maintaining classical ideas has to be carefully considered. In
this perspective, an important research question for our work can be concisely
expressed as follows:
Q6: What are the main criteria to be respected for producing interactive simu-
lations using the sum over paths approach which are effective in sustaining
students’ understanding, but at the same time do not mislead them into
attributing classical properties to the quantum object?
Experimental activities to be performed with students are of enormous value
in educational practice. In the teaching of quantum physics, adding an exper-
imental dimension to the learning path is not a trivial task, as in many cases
expensive and cumbersome apparatuses are required, which are not a viable
option in school laboratories. Where possible, teachers can rely on university
structures; however this still does not solve completely the problem, since even
in this case, normally only one apparatus of a kind is available, which can be
an obstacle in organizing group activities, and often forces to limit to teacher-
centered “demonstrations”, which are much less effective. Thus, experimental
1
The word anschaulichkeit in German means “visualizabiity”, but its semantic extension
also includes the concept of “intuibility” [31].

4
1.3. Dissertation outline

activities using low cost, and possibly easily transferable materials, are of high
value to teachers. Our work in this area can be seen as an attempt to provide
an answer to the following research question:

Q7: Can we sustain the work of teachers by designing low cost experiments
which are meaningful for the understanding of quantum physics?

It is not sufficient that teachers are formed on the basics of quantum physics
using an innovative approach, to provide them with enough motivation to in-
troduce innovation in educational practice. And correspondingly, the achieve-
ment by students of a good understanding of the basics of the theory has to
be considered a necessary, but not sufficient objective: the epistemological
structure of quantum physics should be accepted by students, and the content
should be made relevant to them. In the recent literature, these concepts have
been framed in terms of appropriation [1, 32, 34, 35]. Without entering, in the
context of this introduction, in a detailed analysis of the concept of appropri-
ation, which will be performed in Section 5.1.5, we articulate here two further
research questions, condensing our objectives in this area.

Q8: In the context of compact teacher training courses, is making the con-
ceptual and epistemological structure of quantum theory transparent and
easily verbalizable, through the Feynman approach, a viable strategy for
initiating appropriation, and providing teachers with the instruments and
motivation for proceeding in the study, and pursuing innovation in educa-
tional practice?

Q9: In the high school context, can a teaching-learning sequence based on

Feynman’s formulation form the core nucleus of a complex, multi dimen-
sional learning environment meant to encourage appropriation of the sub-
ject matter by students? In this respect, is a comparison with other ap-
proaches to the teaching of quantum physics possible?

The nine research questions herein formulated represent “longitudinal” threads

running throughout this dissertation, and can be considered as the areas in
which original contributions, sometimes very small, have been brought to the
scientific debate on the teaching of quantum physics from our, collective as a
group and individual, research work.

1.3 Dissertation outline

This dissertation is organized as follows: in Chapter 2 we provide a general
overview of our reference framework in physics education, reviewing the works
and approaches of those authors which we consider as milestones for designing,
organizing and interpreting the results of our work. In Chapter 3 we provide a
traditional review of the existing literature on the teaching of quantum physics,
focusing more, although not exclusively, on works dealing with high school or

5
 1. Introduction

non-specialist undergraduate students, or with the education and formation

of teachers. Chapter 4 contains a global account of Feynman’s approach. We
discuss theoretical aspects (the time dependent path integral formulation, and
the sum over paths representation of the energy dependent Green function)
and review in some detail most of the educational proposals which have been
advanced, based on Feynman’s perspective. In Chapter 5 we provide a chrono-
logical account of the design process of our teaching-learning sequence: we
start from the initial design choices and their rationale, we proceed with re-
porting the two successive versions which we used for teacher training in 2014
and 2015, and finally we account for the process of revision and expansion
which led to the 2015 experimentation in high school. In Chapter 6 we discuss
the results of our experimentations, both in the quantitative and qualitative
dimensions. Finally, in Chapter 7 we draw the conclusions for our work, we
evaluate the tentative, provisional answers that we could give to the research
questions enumerated in the previous section, and we discuss some ideas and
perspectives for future research directions.

6
Chapter 2
Reference guidelines from
physics education research

In this chapter, we will provide a picture of our general reference framework in

physics education; in particular, we will discuss those theoretical paradigms,
approaches, and methods of data analysis which have been incorporated as
essential elements into our work.

2.1 The Pavia approach to physics education

Research in physics education started in Pavia, as in a handful of other Uni-
versities in Italy, in the 1970s [36] when a small group of researchers entered
the vast international debate on the teaching of science at all school levels.
Most of the capital of research methods and approaches accumulated by the
Pavia group in 40 years has entered, in one way or another, in the process
of designing and testing of our learning path on quantum physics. In the fol-
lowing sections we will provide an overview only of those research directions
which were more directly and explicitly relevant to this work, starting from
the approach of Pavia to the design of teaching learning sequences.

2.1.1 The design of teaching-learning sequences

A very important line of inquiry in physics education, some aspects of which
date back to the early 1980s, concerns the design and implementation not of
long-term curricula, but of sequences for teaching focused on a specific topic
[16, 17]. A distinctive feature of such research area is the dual character of its
products, which involve both a priori theoretical research and development,
and a close connection and interplay with the practical activity of teaching and
learning a particular topic. Although various names have been given to such
type of research products in the past, the term “teaching – learning sequences”
(TLS) has become almost universally recognized starting from the beginning

7
 2. Reference guidelines from physics education research

of the XXI century [37]. Closely related to the design and validation of TLS
is the model of “educational reconstruction” [38] which holds that a clarifica-
tion of the subject matter is a key issue to be considered, in preparation to
the design of instructional material on a certain content. This is a process
called “elementarization”, which leads to the reconstruction of the core ideas
to be taught, taking into consideration also the ideas developed by students.
Although the original model by Kattmann [38] was strongly connected to a
constructivist epistemological position, it has been interpreted by several re-
searchers [39], including the Pavia group, in a way less tightly connected to a
particular perspective on the nature of science.
A distinguishing feature of the research in Pavia on the design and validation
of TLS is a special attention to teachers, which means taking in consideration
both their motivations and their needs. In the Pavia approach, which is com-
pactly presented in Ref. [40], a basic assumption, which has been formulated
also upon reflection on the long experience in teacher training courses, is that
teachers can only be productively engaged in renewing instructional methods
and implementing research based proposals, if they feel the need of enlarging
their own view and understanding of the topic. In other words, the primary
motivation for a teacher to innovate his educational approach, is not pedagog-
ical or didactic, but cognitive and content related. In order to stimulate such
urge in teachers, the process of implementation of TLS includes a preliminary
work with teachers, in which they are encouraged to critical reflection on the
physical content, its cognitive concatenations, and its educational transposi-
tion [41].
Concerning the attention to the needs of teachers, in the Pavia tradition TLS
are designed with an “open source structure” which is intended to facilitate the
reproducibility of the sequence in an actual classroom context. The structure is
made of a core backbone of contents, conceptual correlations, and fundamen-
tal methodological choices; and a cloud which includes peripheral elements,
content which can be re-designed or omitted, or in which new content can be
added by the teacher. Elements which are redesigned or added by the teacher
can enter the final research product, in an interplay which is inspired by the
tradition of action research.
In the preliminary process of educational reconstruction of the content, three
main moments or dimensions are identified [40]:

1. Analysis of didactic research concerning conceptions developed by students,

and previous TLS on the topic;

2. Overview of the usual treatments, including textbooks and common teach-

ing practices, as well as feedback from preliminary testing of materials with
small groups of students and teachers;

3. Critical analysis of the scientific subject, including its historical develop-

ment, its epistemological structure, and its practical applications.

8
2.1. The Pavia approach to physics education

Overall, the contribution of Pavia to research on the design and validation of

TLS is an important step in bridging the gap between research projects and
actual classroom practice [41]. Based on the above principles, the group has
proposed to the scientific community TLS on the specific subjects of friction
[40]; free fall motion [42]; the greenhouse effect [43]; rolling motion [44] and
magnetic forces [45, 46]. The activities carried out during the study and exper-
imentation of these different projects allowed the group to construct a general
methodological approach for working with teachers, which includes discussion
of innovative ideas, insights on specific content and experimental aspects, con-
tinued interaction in the phase of practical implementation in class, discussion
of the results obtained with students, reformulation and readjustment of the
original proposal [36].

2.1.2 Innovation in teaching modern physics: the case

of relativity
Between 2006 and 2010, De Ambrosis of the University of Pavia in collabo-
ration with Levrini of the University of Bologna coordinated a project on the
formation of secondary school teachers on the teaching of special relativity
[1, 47] in the context of the IDIFO Master on “Innovation in physics education
and vocational guidance” [48]. In the course, which was entirely conducted
online, they analyzed and discussed with teachers the innovative proposal by
Taylor and Wheeler [49] which approaches relativity from a geometric point of
view, building a language and tools potentially fruitful for understanding gen-
eral relativity. The proposal was compared to the approach by Resnick [50],
which is the main reference for teachers and authors of textbooks. The project
represented the attempt of getting teachers involved in educational innovation
through a path of progressive appropriation, which included analysis of the two
different proposals, reflection on their historical grounding and epistemological
structure, and the design of possible sequences for the practical implementa-
tion in class.
In the development of this project, a further capital of experience in the for-
mation of teachers was accumulated, dealing specifically with the problems
raised by modern physics, which was of invaluable help for the work discussed
in this dissertation. In fact, there are several points of contact between Taylor
and Wheeler’s proposal for special relativity and our own approach to teaching
quantum physics (which, as we will see in Chapter 4 comes from a research
line which was initiated by Taylor himself): in both cases, a non-historical
reconstruction of the theory is proposed, which is innovative with respect to
the perspective usually adopted in textbooks, and has the explicit aim of high-
lighting and making transparent the conceptual aspects that have proved most
fruitful in the subsequent development of physics. Not surprisingly, the two
proposals encounter similar initial resistances from teachers who, often not
being completely comfortable even with the content of textbooks, are intimi-

9
 2. Reference guidelines from physics education research

dated by innovation. The appropriation path developed by De Ambrosis and

Levrini in order to overcome such initial resistances, and raise the awareness of
teachers on the necessity of educational innovation, has been one of the main
reference points for the present work.

2.1.3 The use of ICT in education

The research group in Pavia has accumulated in the years an enormous capital
of experience and expertise in integrating education with information technol-
ogy, in a history whose origins can be traced back to the very first days of the
computer age. We only recall here the two main directions which are pursued
to this day, and which were more directly relevant for our work:

ˆ Research on computer simulations, and their use in the context of labora-

tory experiments [51, 52],and inquiry and discovery activities [44, 53]. Our
decision of assigning a central role to to GeoGebra simulations in this work
was firmly grounded on the previous work of the group in this area.

ˆ Research on the use of ICT as laboratory instruments, until a few years ago
concerning mainly sensors and MBL (Microcomputer Based Laboratories)
[54, 55], nowadays more focused on the acquisition and analysis of videos and
photos, through cameras and cell phones, and their analysis using software
tools such as Tracker [56, 57, 58]. The experimental activities included in
our learning path [59] were developed using the expertise accumulated in
this research line.

2.1.4 The role of history of physics

An important theme of the research conducted in Pavia is the connection
between history of physics and education, which started from the works of
Bonera on Galileo and Volta [60]. Such thread, which is also a major area
of international research [61], was often stimulated by the tight connection
with the group of history of physics. Different themes were explored, such as
the one on the concept of energy [62, 63], and the one concerning historical
experiments [58, 64, 65, 66], both of which continue to be pursued to this day.
The experience in this field has been especially relevant in the process of re-
thinking the sequence, expanding its historical-epistemological dimension, in
preparation for the first practical implementation in class (see Section 5.5).
For example, the works of Besson on the different types of explanations used
in physics, both in education and history [67]; and of Giuliani [68, 69] on the
philosophical positions of Hertz and Maxwell were taken into account in the
design of the initial questionnaire (Appendix A.3.1).

10
2.2. The concept of appropriation

2.2 The concept of appropriation

Appropriation is a term whose semantic extension in the social sciences is

rather vast, as it is not always used with similar meanings in different fields,
and in some cases, such as in anthropology and ethnography, it is impregnated
with negative undertones. In science education, however, “appropriation” is
usually intended with a positive connotation, as a process in which the learner
internalizes the acquired content, incorporates it into his own discourse, and
is then able to use it in a personal way, expressing it with a language and
terms not uncritically retrieved from external authorities, for interacting with
others. The term has been used in this way for example in Ref.[70], and in the
context of the project on the teaching of relativity coordinated by De Ambrosis
and Levrini [1, 47] which was discussed in Sec. 2.1.2. Recently, Levrini and
co-workers have attempted to define the term more precisely, transforming it
into a full-fledged theoretical construct, and at the same time to operationalize
it [34], providing criteria, or “markers” which could be used to check whether
a process of appropriation, for a particular learner, had or had not occurred.
In doing so, and referring specifically to the case of appropriation in young
students, such as those in high school, Levrini and co-workers somewhat ex-
pand the scope of the term, by arguing that appropriation can be seen as the
process by which disciplinary content becomes so relevant to the individual,
to be chosen as an element in the construction of his personal narrative of the
self, i.e. his identity. Although their work on precisely defining the contours of
this new nuance of the concept of appropriation is not yet finished, we found
the idea extremely interesting and worthwhile.
The idea of favoring appropriation has entered in the design of our learning
path in quantum physics from the very first day, since one of the cornerstones
of our work was the long time experience of De Ambrosis in the formation
of teachers on relativity. In fact, perhaps the primary goal of our work, and
the main reason underlying the choice of Feynman’s approach, has been to
make the conceptual and epistemological structure of quantum physics fully
visible and verbalizable, developing an adequate language for discussing it with
students. In the process of rethinking our teaching learning sequence for an
implementation in class, we decided to expand our objectives, and take up the
challenge of favoring and verifying students’ appropriation, also in the wider
sense described above.
The cornerstones of the work in Ref. [34] are the definitions of a set of guide-
lines for learning environments designed to foster appropriation; and of a set
of operative markers for verifying the successful occurrence of appropriation in
students. Concerning the construction of “properly complex” [35] learning en-
vironments, Levrini and co-workers, by combining and re-shaping suggestions
contained in the literature [71, 72, 73, 74], identify the three following criteria

11
 2. Reference guidelines from physics education research

1
:
1. Longitudinality expresses the idea that the learning of physics is a contin-
uous process of widening and refining acquired knowledge, which must be
globally coherent. Even if teaching is approached in self contained units
or sequences, the connections with models, phenomena and theories pre-
viously studied must be systematically presented, and, from the point of
view of language, if some terms or concepts undergo shifts in meaning in
the passage from one theory to another, the differences and analogies must
be explicitly discussed.

2. Multi-dimensionality is intended as the requirement that the physical con-

tent is analyzed and compared according to different dimensions; thus, not
only from a conceptual, experimental, and formal point of view, but also
considering their philosophical-epistemological implications.

3. Multi-perspectiveness means that the content is analyzed under two differ-

ent perspectives which are both internal to the boundaries of the discipline,
i.e. two different physical perspectives or approaches. For example, in the
context of their study on thermodynamics [34] the authors systematically
compared the macroscopic and microscopic approaches. As we will see
in Section 5.5, in the context of quantum physics is not easy, nor per-
haps necessary, to precisely draw borders between the concepts of multi-
perspectiveness and multi-dimensionality.
The design of a learning environment respecting such criteria is not a trivial
task; it involves producing teaching materials which are relevant from a cul-
tural, and not only scientific, point of view; and introducing activities, such
as open discussions concerning different perspectives or epistemological views,
which challenge the authoritative image of science, in which a unique point
of view is legitimate [75]. Operatively verifying whether appropriation has
successfully occurred in individual students requires letting them speak freely
about the content, possibly in the absence of the teacher, and with less con-
straints than usual; this is best realized through individual interviews con-
ducted by other researchers than the teacher. The recorded interview is a
crucial piece of data in this work, in fact it must be analyzed in depth, of-
ten listening to it several times, in search of revelatory signs of connections
between the disciplinary content acquired, and the student’s individual narra-
tive of the self. Such signs may be in the form of recurring terms or phrases,
linguistic idiosyncrasies, or a real thesis explicitly brought to surface during
the interview. Once identified, also by comparison with data from previous
activities, such elements are organized in profiles, describing how the student’s
discourse seems to be organized around his idiosyncratic terms, expressions,
or his thesis. Finally, all the data are systematically re-evaluated, checking
1
We reorder the criteria with respect to their presentation in Ref. [34] according to our
own perception of their relative importance.

12
2.3. Conceptual change in the perspective of framework theories

whether a “signature” idea clearly emerges from the student’s profile and data,
which respects five operational markers:

a. it is authentic, in the sense that it is recognizable as personal, and it is ver-

balized using terms and expressions which are not borrowed from external
authorities, such as the teacher or the textbook.

b. It is grounded in the discipline, in the sense that it respects the disciplinary

norms of physics, and it used by the student for coordinating the physical
content in a way which is meaningful to him, and scientifically valid.

c. It is thick, meaning that it is grounded in, and inseparable from, the stu-
dents’ metacognitive and epistemological discourse.

d. It is non incidental: the development of the idea is connected to the stu-

dents’ individual story and intellectual trajectory, and as such it does not
appear only in one isolated episode (for example, in the interview only),
but it can be traced, maybe in an embryonic form, in different classroom
activities.

e. It is bearer of social relations, i.e. the idiosyncratic idea identifies a role

or position of the student within the class community, and conversely, the
development of the idea is not separable from the overall dynamics of the
class.

In our own work (Section 6.3.5) we chose six students for the interviews. In
the preliminary data analysis contained in this dissertation, we identified one
clear case of appropriation, and one which is dubious, but which we are more
prone to consider as one of non appropriation. Our work so far confirms that
the five markers introduced in [34] are extremely relevant and operationally
effective, although marker (e.) is often the most difficult to interpret for indi-
vidual students.

2.3 Conceptual change in the perspective of

framework theories
Several authors have considered the problem of teaching quantum physics from
the perspective of conceptual change [25, 76, 77, 78]. The so called “classi-
cal approach” to conceptual change [26, 27] can be defined as an attempt at
transposing the Kuhnian theory on scientific revolution [79] to science educa-
tion. Its basic principles are summarized in the dissatisfaction-intelligibility-
plausibility-fruitfulness model by Posner and co-workers:

ˆ Dissatisfaction. The learner must first realize that his existing conceptions
cannot explain the new evidence at hand, and that a radical change is
necessary. This result will not be achieved with a single anomaly, but a

13
 2. Reference guidelines from physics education research

whole range of problems which are unsolvable with the old approach mist
be collected before the learner accepts the need to undertake conceptual
change.

ˆ Intelligibility. For a learner to accept and accommodate a new conception,

he must find it intelligible. The model should make sense to him, and he
should be able to explain that concept to other students. Initially, analogies
and metaphors can help making the new model intelligible.

ˆ Plausibility. The new conception must be plausible for it to be accommo-

dated; it must not appear to be only an ad hoc construction made to deal
with anomalous evidence, but must also be consistent with the previous
model on facts that it could explain.

ˆ Fruitfulness. The learner must find that the new model has the potential
to be extended to other incidences and open up new areas of inquiry.
Posner’s conceptual change model gave rise to the “misconceptions” movement
[80, 81, 82] arguing that the learning of science involves the replacement of
persistent, theory-like erroneous conceptions. According to these authors, such
replacement process should involve a global restructuring, in analogy to how,
in Kuhn’s theory, a new paradigm completely replaces the previously dominant
one.
In contrast, Vosniadou in her “framework theories” approach to conceptual
change [26, 27, 28, 83] argues that a global restructuring, meant as a complete
replacement of a theoretical structure with another, in a more or less con-
centrated interval of time, seldom occurs in instruction; and in some specific
domains of science, it never does. More often, the initial framework theory
and the one which is being acquired through instruction2 may coexist in the
mind of the learner for a long time, forming a dynamical system in constant
development. During the process of acquiring information incompatible with
the pre-existing framework, the learner may develop internal inconsistencies,
and reorganize his knowledge in “synthetic” conceptions or models, which are
forms of hybridization between the two frameworks.
One important difference with the misconceptions movement is that, accord-
ing to Vosniadou, initial framework theories are constantly evolving under the
influence of the learner’s experience, and do not contain crystallized miscon-
ceptions; actually, such theories are usually internally consistent and have,
within their domain of validities, a predictive and explanatory value. Thus, a
misconception is more often an inconsistency arising from the conflict between
the initial and acquired frameworks, and in this sense is in itself a form of
hybridization.
Similarly to the authors of Ref. [76] we believe that Vosniadou’s framework
theory approach is relevant in student instruction on quantum physics. For
2
These theories may be understood as a naı̈ve and a scientifically valid theory, but the
definition also applies to the two frameworks of classical and quantum physics [76].

14
2.4. Quantitative data analysis using knowledge integration rubrics

example, in our teaching experience we observed several cases of students spon-

taneously producing hybrid, synthetic models; many others are reported in the
literature [20, 21, 22, 23, 24, 84, 85]. We also found other signs of a continued,
complex interaction between two coexisting frameworks (see Section 5.3.5). It
is important to note, however, that the formation of hybrid conceptions which
incorporate elements of a previously existing model and the one which is being
acquired through instruction is recognized and explained also by authors who
do not adopt the theoretical perspective of framework theory [86, 87], for ex-
ample in studies advancing the research line originated from diSessa’s theory
of p-primes [88, 89, 90, 91].
Vosniadou and co-workers provide several suggestions for designing learning
environments able to promote successful conceptual change from the perspec-
tive of framework theories [27, 83]. Here we report three of these suggestions,
which have more directly influenced the design of our teaching -learning se-
quence (see Sections 5.1.2 and 5.1.1):

1. Explain, rather than replace. Instruction should not simply tell to learners
that their existing ideas are wrong and should be replaced, but focus on ex-
plaining how the new framework can be consistent with their initial models
and explanations3 .

2. Facilitate meta-conceptual awareness, or in other words try to make students

explicitly aware of the structure of their initial explanatory frameworks.
Give students time for discussing their conceptions, and the possible sources
of conflict with the new model. Make the conceptual structure of the new
theory transparent, explicitly highlighting those ontological categories that
must be created anew.

3. Carefully consider the order in which the material is presented; try to predict
at what points in the learning process contradictions may be produced,
synthetic conceptions may be formed, and develop strategies to counter
them. At this aim, use cognitive conflict on a local scale. Make the new
model more easily available to the student by developing reasoning and
procedural skills, and performing hypothesis - testing activities.

2.4 Quantitative data analysis using knowledge

integration rubrics
In our work we often used open response questions as tools to collect data on
student learning. This type of items is recommended in the literature for the
evaluation of higher level skills and conceptual learning [92, 93, 94], and also
allows to collect other useful information, such as data on the language used by
3
This suggestion, in our context and in the language of the previous section, can be
understood in terms of longitudinality

15
 2. Reference guidelines from physics education research

students. Open response items, especially if the space available for the answer
is not restricted, also give students the possibility of arguing in favor of their
thesis or explanation using supporting evidence, an activity which is similar
to scientific reasoning [95, 96]. Questions in this category require more work
than multiple choice items, especially if a quantitative analysis is desired: in
this case, a clearly defined and fully elaborated scoring rubric to categorize
and classify students’ answers must be developed. The scoring rubric is as im-
portant as the items themselves, because it explicitly fixes the criteria used to
evaluate student’s responses, and also determines the relative weights assigned
to each of the criteria, since these are reflected in scores. Also, using a rubric
reduces the possibility of non uniformity of evaluation, which is especially use-
ful if different researchers examine parts of the same data set [97].
Several of the open response items we used can be classified as “explanation
items” [98, 99] Such questions require to explain a principle or phenomenon
according to the norms of physics, and making scientifically valid connections
with other phenomena, or different ways to understand the same principle.
In the literature, a semi-standardized way to create scoring systems for such
types of questions has been developed, called “knowledge integration rubrics”
(or often KI rubrics.).
Knowledge integration is a theory that represents cognition in terms of “multi-
ple, diverse, and sometimes contradictory ideas students have about scientific
phenomena, and links they make among these ideas” [100, 101]. From the
knowledge integration perspective, science learning occurs when students are
solicited to articulate and verbalize their ideas about the curriculum topic, add
new normative ideas to their repertory, develop scientific criteria to distinguish
between ideas, and form a more coherent view of science as a result of inte-
grating various scientific ideas [102, 103]. A signature mark of the knowledge
integration perspective is the emphasis on the importance of internal coherence
in students’ scientific frameworks, and in the connections among their ideas
[101].
We do not here enter into the details of the knowledge integration paradigm;
but the basic premise of “assessing integrated science understanding” [104] on
which KI rubrics are based is certainly widely agreeable. In fact, KI rubrics are
used as an efficient instrument for evaluating successful learning also by au-
thors who do not adopt in full the knowledge integration perspective [105, 106].
Referring to physics in particular, “assessing integrated understanding” can be
translated into the idea that a key element to consider in evaluating students’
learning is their ability to connect different physical concepts or phenomena in
a scientifically valid way for argumenting an explanation.
In our study we used five levels knowledge integration constructs [107, 108],
with a general structure as the one exemplified in table 2.1:

In the table, the precise meaning of a “link” is left intentionally vague; in

individual KI rubrics (see Chapter 6)it has to be specified which principles,
phenomena, or sometimes formal computations are considered valid “links” for

16
2.4. Quantitative data analysis using knowledge integration rubrics

Table 2.1: General form of a five level KI rubric for explanation items as the
ones used in Chapter 6.

Score Level Description

Does not provide an explanation, or the explanation

0 Irrelevant
is unrelated to the question.
1 Non-normative The explanation is wrong or non normative.
The explanation appeals to a correct normative prin-
2 Partial
ciple but contains no scientifically valid links.
The explanation is correct and contains one scientifi-
3 Full
cally valid link.
The explanation is correct and contains more than
4 Complex
one scientifically valid link.

a given explanation item.

17
2. Reference guidelines from physics education research

18
Chapter 3
A panorama of research on
teaching quantum physics

In this chapter, we offer a wide angle view of the large part of the existing
literature on the teaching of quantum physics which we consider relevant for
our dissertation. Premising that, in any case, the review has no pretense
of being exhaustive of all works in the field, some general criteria which we
adopted are as follows.

ˆ In the part of the review concerning students’ conceptual difficulties and

mental models (Sec. 3.1.1) we only took in considerations works dealing
with subjects which are covered in our sequence in some significant way.
Students’ difficulties on points which at the present stage are not included
in our learning paths, or only touched in a very limited way, such as spin or
time evolution, were not included in the review. However, for those concep-
tual difficulties which we chose to review, works reporting their appearance
at all levels of instruction were included.

ˆ In the subsection concerning student’s difficulties with classical concepts or

mathematical formalism (Sec. 3.1.2), in addition to the above criterion, we
also limited our attention to studies referring to courses which were algebra-
based, or included only high school level calculus. Students’ difficulties
with advanced mathematical and classical concepts (e.g. Hamiltonians) in
courses based on operators on Hilbert spaces are well studied, but in general
were not considered relevant for our work at the present time.

ˆ Section 3.2 and its subsections, in which we discuss approaches, proposals,

and tools for the teaching of quantum physics at an elementary level, can
be considered a “review” only in a very broad sense. In fact, in this part
we dropped all preoccupations of being exhaustive, and the section must be
regarded as a presentation of the main research lines and directions, citing
some relevant works.

19
 3. A panorama of research on teaching quantum physics

In writing this chapter, we often consulted secondary sources such as published

reviews [109, 110] and Ph. D. dissertations [111, 112, 113, 114, 115, 116, 117].
For the part of the review concerning students’ models we are particularly in-
debted to the extensive review by Falk [110, 114]; while for the section on the
main instructional approaches to the teaching of quantum physics at elemen-
tary level we followed in its general lines the classification proposed by Stefanel
[15].

3.1 Research on student difficulties

3.1.1 Conceptual difficulties and students’ models
Many difficulties in learning quantum physics can be classified as conceptual,
in the sense that they are related to students’ intuitive, qualitative reasoning
schemes and mental models. Difficulties of this kind often persist even when
students have successfully acquired procedural abilities, so that in some cases
they may be able to solve traditional problems requiring mathematical compu-
tations, but not to answer to basic, qualitative questions on the behaviour of
quantum objects and systems. For example, Ireson [118] finds that differences
in the answers of pre-university students to conceptual items before and af-
ter instruction are scarcely statistically significant. In a study concerning the
same level of instruction, Papaphotis and Tsaparlis [119] report that students’
performance in the algorithmic and conceptual dimensions are uncorrelated.
Arons [120] reports that even graduate students in physics have difficulties
in providing qualitative explanations for the basic structure of atomic spec-
tra without recurring to detailed calculations. The authors of the Quantum
Physics Conceptual Survey (QPCS) found that the most difficult questions in
the repository they developed turned out to be those requiring an interpre-
tation or qualitative explanation of physical phenomena [121]. Other studies
[84, 122, 123, 124] report much poorer performance on conceptual questions
than on those requiring algorithmic competence. According to all these data,
students can grasp how to use the computational tools of quantum mechanics,
even without possessing a consistent mental model to make sense of them [125].
As an extreme consequence, the performing of computations may turn into a
meaningless activity [124]. On the other hand, conceptual understanding of
quantum theory, once acquired, appears to be retained for a long time [126].
Empirical research on the nature of students’ mental models about conceptual
issues in quantum physics started in the 1980’s [114] as a means to better un-
derstand the learning process, and support educational innovation. This is part
of a broad research line in physics education, and not related only to quantum
physics. Depending on the perspectives of authors, student’s models contrast-
ing with the currently accepted scientific view may be called also “misconcep-
tions”, “alternative conceptions”, “inappropriate conceptions”, or “depictions”,
which is a terms used by Falk [114, 110] to avoid hinting to a particular view

20
3.1. Research on student difficulties

on cognition. We will use some of these terms interchangeably in the following

discussion, however we will generally refrain from using the word “misconcep-
tion” as it is too heavily charged with meaning referring to a particular point
of view on the learning process (see Sec. 2.3). A recurring characteristic of
alternative models produced by students on different conceptual issues is that
they appear to be constructed by trying to progressively incorporate quantum
elements upon a pre-existing classical framework. Hybrid conceptions con-
structed in this way are only partially consistent, and often fragmented, but
retain essential features of the classical world view. The end result may be
that students have difficulties in distinguishing between classical and quantum
physics, and cannot name any sharp differences between them [127].

3.1.1.1 Wave particle duality

Within the usual picture of the Schrödinger formulation and Copenhagen in-
terpretation, “wave particle duality” can be understood as the property of the
wave function of “collapsing” to a highly localized form due to a position mea-
surement; in this way, the quantum object can be said to (temporarily) acquire
particle-like properties. The concept of wave particle duality is well known to
be problematic in the teaching of quantum physics, as it has been in its histor-
ical development, and students’ models of wave particle duality are extremely
well studied in the literature.
Before the beginning of their instruction, students typically perceive light as a
wave, and matter as composed of pointlike particles, such as electrons. When
first exposed to quantum physics, students usually understand to some de-
gree that a purely wave-like or purely particle-like description is not adequate,
but this may happen in a non uniform way. According to several authors
[77, 124, 128, 129], students may attribute some sort of dual nature to pho-
tons, while continuing to model electrons as purely classical particles. For
example, in Ref. [124] a sample of n = 236 final year high school students
is studied; it is shown that 90% of students attributes some form of dual na-
ture to photons, but only 41% to electrons; in fact, 59% of students interprets
electrons as being simply particles. This evidence has lead some authors to
suggest starting instruction on quantum concepts from electrons, rather than
photons, since the initial classical model of the electron which students possess
seems more stable and less prone to trivial hybridizations [23, 130]. However,
not all studies confirmed this difference [131]. Also, in the context of the two
slit experiment, students often form the synthetic conception that interference
is due to particle-particle interactions [132], and electrons have been found to
more strongly evocate this model [77] partly because of their charged nature
[133].
Dualistic models constructed by students in order to accommodate new quan-
tum phenomena into their existing classical frameworks [128] are often in-
consistent, fragmented [134], and may be coherent only in the context of an
individual experiment [133]. The particle ontology is often dominant in these

21
 3. A panorama of research on teaching quantum physics

productions [131], and a common alternative conception consists in depicting

classical particles as following sinusoidal trajectories [22, 132, 78, 135, 136, 137]
or undergoing random oscillations reminiscent of Brownian motion [131]. In
Ref. [137], 17% of n = 550 German pre-university students depicted the pho-
ton as moving in wave-like trajectories. When the wave ontology is dominant,
students’ hybrid models may be of the matter-wave [77, 111, 133], or “energy
lump” type [128, 129]. The formation of these models is not limited to high
school students or to the very early stages of instruction in quantum theory; for
example in Ref. [77], 20 out of n = 25 second year physics undergraduates pro-
pose a mixed quantum-classical description. In fact, studies have reported that
these models are resistant to instruction, and very little advancement may take
place in conceptual understanding of wave particle duality while progressing
with the study of quantum physics [23, 118, 131]. Even students who get good
grades according to the usual evaluation methods may show poor reasoning
on the duality concept [134]. There are, however, also studies reporting that
students’ ability of producing correct mental models of wave particle duality
does improve with undergraduate college instruction [138].
Difficulties in linking the dual nature of quantum objects with relevant exper-
iments have also been reported [133, 139]. Partly based on these data, some
researchers have suggested that dualistic descriptions be entirely avoided in
education [23, 140, 141].
Some of students’ difficulties are related specifically to the concept of De
Broglie wavelength [124, 142]. A large study carried out by McDermott and
co-workers [142] involved, among other groups, n = 195 first to third year sci-
ence undergraduates who attended an introductory quantum physics course.
Of these, 75% could not consistently relate the wavelength of a massive par-
ticle to its momentum, often because they incorrectly extended mathematical
relations which are only valid for photons, or because they believed the wave-
length to be an intrinsic property of the particle, depending on mass but not
on velocity. The percentage of students showing difficulties with the concept
dropped to 35% after a specific tutorial.
The issue of what models students build for wave particle duality is intertwined
with the questions of which formulation of quantum physics should be adopted,
and which interpretative and foundational attitude, if any, should be taken by
instructors [24, 32, 111, 125, 143, 144, 145]. Many of these perspectives are
contained in works of theoretical education research, which will be discussed
in Sec. 3.2.2. In Refs. [111, 125] the authors perform a detailed comparison of
introductory courses in quantum physics held by instructors choosing different
attitudes towards interpretative issues. One of their main results is that if the
teacher chooses not to take an explicit position on foundational matters, stu-
dents are more likely to remain attached to a classical realist position on several
issues, including wave particle duality. In Ref. [24] the authors adopt the per-
spective of epistemological innovation developed by Bunge and Lévy-Leblond
[146, 147, 148, 149, 150], reporting satisfactory results on the models used by

22
3.1. Research on student difficulties

students to represent quantum objects. Fanaro and co-workers [10, 145, 151]
use the sum over paths formulation to lead high school students to an anal-
ysis of the two slit experiment with electrons. They report that students are
satisfied by the explanation provided by Feynman’s perspective; however, they
also mention that students have difficulties in accepting that electrons, in the
end, do not take only one of the possible paths. Our own results, which have
been reported in [152, 153, 154, 155] and are more extensively presented in
Chapter 6, confirm that students can build satisfying mental models of wave
particle duality, accepting them at least to the extent of considering them
reliable sources for explanations. Whether they accept them as a definitive
description of reality is, as we will briefly discuss in the conclusions, a more
definitive issue to settle.

3.1.1.2 Atomic models

Students’ difficulties and depictions concerning atomic models are also ex-
tremely well studied. In some cases, findings of these studies overlap with
those presented in the previous section, for example when discussing students’
representations of electrons around the nucleus. A number of researchers have
reported that students have difficulties in abandoning the simple planetary
model of the atom. For pre-university students who have received only intro-
ductory instruction about quantum physics, this has been confirmed by studies
conducted in several countries [23, 84, 85, 124, 128, 137, 156, 157, 158]. For
example, Papaphotis and Tsaparlis [84] perform a detailed study with n = 125
Greek high school students who had been taught basic quantum concepts and
the orbital model of the atom, reporting 72% of answers classifiable as “de-
terministic” to the question “Make a drawing depicting the hydrogen atom
as you imagine it is in reality.” Notably, they also report that students pro-
duced several variants of the planetary model, including synthetic conceptions
they invented on their own: these variants include point-like electrons moving
in sinusoidal-shaped trajectories oscillating around the classical orbit, going
in elliptical orbits of variable inclination, or discontinuously jumping around
the circular orbit; and electrons depicted as small wave-like “clouds”, also cir-
cling around the nucleus. The same basic kinds of hybrid models are also
reported in Ref. [85], where in a sample of n = 142 German pre-university
students, approximately 50% represents the atom using a variant of the plane-
tary model. Olsen ([124, 159]) in a study with a cohort of n = 236 Norwegian
students in the final year of high school, using a multiple choice question in
which different models of the atom are represented, finds 51% of deterministic
answers, consisting, in order of preference, of the electron staying in one of
several concentric spherical shells (31%); moving on rotating-axis elliptical or-
bits (16%), moving on planar circular Bohr orbits (3%) or on wave-like orbits
(1%). The same issues have been reported, although with lesser percentages
of students remaining stuck to the planetary model, for pre-service teachers
[137, 158, 160, 161, 162]; for chemistry [163] and physics [164] undergraduates,

23
 3. A panorama of research on teaching quantum physics

and in the context of an introductory chemistry course for all science under-
graduates [115].
The educational use of the Bohr model has been criticized by several authors,
as they have reported that it offers to students a powerful explanatory image
which is very close to their classical conceptions, with the consequence that
they may be reluctant to go beyond it [25, 23, 76, 118, 135, 165, 166]. Some
of these authors have suggested to avoid it completely in introductory courses.
However, suggestions on how to substitute it differ, and empirical results are
still scarce. In Ref. [23] a two group pre- and post-test study is performed
with n = 270 total pre-university students. Prior to instruction on orbital
theory, 63% of students used variants of the planetary model to represent the
atom. Then the students were divided in two randomly selected groups. To
the “control group” a traditional course was given, and in the post-test still
60% of them maintained a picture of the atom based on classical orbits. For
the test group, to which a modified course was given, designed to avoid the
Bohr model and all analogies with classical physics, the percentage dropped
to 22%.
The authors of Refs. [165, 168, 169, 170, 171] have proposed a different in-
termediate model than the Bohr one for educational use, in high school but
also at undergraduate level, which they call “Electronium”. The model was
developed in the context of the Karlsruhe Physics Course [172], and is based
on depicting the electron as a fictitious substance or fluid of variable density,
similarly to the De Broglie - Madelung hydrodynamic interpretation [173, 174].
They suggest that such model may be more efficient as a stepping stone to-
wards a full quantum perspective, and report supporting evidence, in the form
of interviews to individual students, and observations on the evolution of their
mental models through a sequence of meta-stable conceptions. This approach
has been criticized in Ref. [164] for introducing a new hybrid model with scarce
connection with the historical development of atomic physics. Other authors
[135, 175, 176] have proposed to concentrate on a spectroscopic approach and
the use of energy level diagrams, rather than spatial descriptions of the atom.
This suggestion is supported by findings reported in the literature [112, 168],
according to which the quantization of energy is readily accepted by students,
and they do not ask for further physical explanation referring to an abstract
quantum model based on energy levels.
However, the question of whether the picture of a planetary model in which
only some orbits are allowed is necessarily a consequence of explicit instruction
has been debated [114, 128]. In fact, students may autonomously form such
model by combining a planetary conception of the atom which they acquired
very early in instruction, or from external sources, with the idea of allowed
energy levels, in the same way as they construct different synthetic models. In
some cases, this may also be a consequence of a misinterpretation of the con-
cept of orbital [177]. Indeed, some research works have shown [137, 178, 179]
that the introduction of the Bohr model is not, per se, an obstacle for students

24
3.1. Research on student difficulties

in acquiring a full quantum perspective, provided its limitations are discussed

in full (which high school textbooks often do not do adequately [180]), and
the connections and differences with other models, including the currently ac-
cepted one, are evaluated with students.
Kalkanis, Hadzizaki and Stavrou [76, 158, 167] while highlighting the edu-
cational shortcomings of the Bohr model, do not agree with the proposal of
avoiding it, but suggest instead to directly juxtapose it to the currently ac-
cepted one, leading students to consider the unacceptable classical elements of
the former, such as its deterministic aspects. In their study, they work with a
composite group of undergraduates, comprising perspective teachers and stu-
dents of history and philosophy of science; both these groups had only studied
elements of the quantum model of the atom in high school. The course pro-
posed by the authors, which also includes the exploration of a simulation of
the orbital model, is given to a test group of n = 102 students. Two weeks
after the end of the course, students fill a questionnaire which is also given to a
control group comprising n = 98 students who received no further instruction
after high school. Results show that 94% of students of the control group rep-
resent the atom using variants of the planetary model, while 99% of students
in the test group provide an acceptable representation of the quantum orbital
model and, according to other items of the questionnaire, more than 90% also
interpret the model correctly.
In Ref. [178], the authors report the results obtained in the context of a one
semester modern physics course for engineering undergraduates covering an
innovative research-based curriculum which they developed. The course in-
cludes several collaborative activities between students, and uses interactive
simulations from the PhET repository ([181, 182]). In the part concerning
models of the atom, those by Bohr, De Broglie, and Schrödinger are discussed
and compared, highlighting connections and inadequacies of the intermediate
model. Data obtained from final questionnaires show that only 13 − 16% of
students remain stuck to represent the atom using the Bohr model only. The
authors conclude that teaching the Bohr model is not an obstacle for students
to learning the currently accepted one.
Besides the spatial model of the atom, student difficulties in interpreting and
connecting atomic spectra and energy level diagrams have also been reported
[112, 115, 120, 122, 176, 183]. In Ref. [115] this kind of difficulties are thor-
oughly investigated in a cohort of n = 65 science undergraduates involved in an
introductory chemistry course; notably,at least 17 of these students (10 more
held mixed or fragmented views) believe that an atom can absorb a photon of
any energy, but the surplus energy that cannot be used for a transition has
some other effect, such as “heating” the atom. The same study and others
also report a difficulty which concerns students thinking that the energies of
emitted and absorbed photons correspond directly to those of the allowed lev-
els, rather than to the difference between the energies of two levels. In Refs.
[176, 183] the authors suggest to deal with this specific difficulty using com-

25
 3. A panorama of research on teaching quantum physics

puter simulations, such as those of the Visual Quantum Mechanics project that
they developed [184].

3.1.1.3 The uncertainty principle

Students’ conceptual difficulties with the uncertainty principle are somewhat
less well studied; and as for the case of wave particle duality, educational prob-
lems are intertwined with epistemological and foundational issues [111, 125,
143]. In its most uncontroversial, widely accepted interpretation, the uncer-
tainty principle is a property of an ensemble of quantum objects prepared
in an identical state; and it asserts that the statistical distributions of mea-
surements of two complementary variables performed on individual elements of
the ensemble will have standard deviations obeying the Heisenberg uncertainty
relation [23, 114]. In other words, if an experimenter wishes to prepare a quan-
tum object in a state in which, for example, its position is well defined within
∆x (in the sense that a measurement of position in an ensemble of systems
prepared identically to the experimenter’s one will give a distribution of results
with standard deviation ∆x); then he must accept that the momentum of the
quantum object is well defined (in the same sense) only within ∆px , with the
product of uncertainties given by Heisenberg’s relation. This statement makes
no hypothesis on the origin of uncertainty, and indeed, its most widely accepted
(but no longer completely uncontroversial) complement is that no further ex-
planation is possible or required, since uncertainty is an intrinsic property of
quantum objects, limiting the information which is in principle available on
them. In more precise, although not so widely used terms, it can be said that
an overwhelming majority of the physicists community nowadays believes that
Heisenberg’s principle is not actually about uncertainty, but about intrinsic
indeterminacy [185, 186].
Basically, two kinds of conceptual difficulties are reported in the literature.
Students may think of the principle as expressing an experimental limitation,
i.e. that due to noise or some other effect, one cannot obtain measurements
which are more accurate than a certain limit [77, 134, 137, 187]. In this case,
often students’ models may also fail to recognize, or underestimate, the sig-
nificance of the principle being formulated in terms of complementary vari-
ables; for example, they may think that neither position nor momentum can
be measured beyond a certain precision [77, 187]. Or, students may think
of uncertainty as caused by measurement, in the sense that its origin lies in a
physical disturbance on the observed system caused by the experimental appa-
ratus [77, 137, 188]. This conception, although not consistent with the current
understanding of the theory, is sometimes defined as “somewhat more accept-
able” [114], since it is connected to the historical interpretation and thought
experiment of Heisenberg’s microscope. It must be recalled that, up to some
years ago, some textbooks still presented the principle based solely on this
explanation [189].
These difficulties have been reported for high school students [118, 170] and

26
3.1. Research on student difficulties

at higher levels of instruction. In Ref. [77] the authors investigate the concep-
tions on the uncertainty principle of a sample of n = 25 Ethiopian second-year
physics undergraduates. They find that only 3 students give to the principle a
proper quantum interpretation: among the other 22, a majority believes that
the principle expresses a limitation of experimental apparatuses which, for a
number of reasons, give a value which does not coincide exactly with the “true”
one, while a minority interprets it as the result of a disturbance on the observed
system by the experimental apparatus.
Müller and Wiesner [137, 190] performed a similar study on a cohort of n = 37
German undergraduate perspective teachers, of which 79% had attended a
course on quantum physics, and 52% had studied elements of quantum theory
in high school. In this case, 15% of students held that uncertainty is unavoid-
able experimental deviation from a true value, while 21% reasoned in terms of
disturbance on the measured system.
Although Heisenberg’s microscope is much less commonly used in education
than the Bohr model of the atom, it could play a similar role in the develop-
ment of students’ conceptions, as an explanatory device which may turn out to
be too powerful to be replaced. In fact, some authors have suggested to avoid
all formulations of the uncertainty principle based on disturbance [23, 118]
while others have proposed to discuss Heisenberg’s model, by juxtaposing it to
the currently accepted view, or in the context of the introduction of historical
debates [32, 189, 191, 192]. In current educational approaches and textbooks,
one of the strategies most often chosen for introducing the uncertainty princi-
ple is through the experiment of diffraction of photons or electrons from a slit
of variable width [152, 193, 194, 195, 196, 197, 198]. In Ref. [199] a positive
influence on students’ mental models is reported by favoring the association
between the uncertainty principle and the ground state energy for a particle
in a box, which is consistent with our own findings.

3.1.1.4 The tunnel effect

Several authors have analyzed students’ conceptual difficulties with quantum
tunneling. In the context of this dissertation, the most relevant difficulties
are those concerning the basic problem of time-independent tunneling from a
square potential barrier, which are also the most well studied. A very com-
monly reported problem in students’ understanding of tunneling is the idea
that a particle loses part of its energy crossing the barrier. In Ref. [200]
this conception is reported for 11 out of n = 15 physics undergraduates who
had completed an introductory level course on quantum physics at Univer-
sity of Maine, and it has been explored in a large number of other studies
[113, 123, 201, 202, 203]. The origin of this conception has been related to
two main factors: on one hand, students may rely on a classical intuition of
tunneling as a dissipative process, similar to a bullet penetrating a solid barrier
[113, 204, 205]. On the other hand, they may confuse the value of the wave-
function, or of the probability density, with the energy of the quantum object;

27
 3. A panorama of research on teaching quantum physics

in doing so, they may also rely on misleading or inaccurate representations of

the wave function, produced by the students themselves, or present in text-
books. For example, authors have highlighted that typical diagrams reported
in textbooks, where potential energy and wave function or probability density
are plotted on the same axes, may play a role [200, 206]. Thus, students may
interpret exponential decay of the wave function in the barrier as a similar
decay in energy [123, 200]. Other kinds of representations of the wave function
which have been associated in the literature to student’s depiction of energy
loss in tunneling, among which the widely recognized “axis shift phenomenon”
[123, 113, 201, 207] are represented in Fig. 3.1, and briefly commented in its
caption.

(a) (b)

Figure 3.1: (a) Inaccurate representation of the tunneling wave function, sim-
ilar to several ones reported in Ref. [113]. The wave function has a shorter
wavelength in the right region. This is related to trying to reproduce the
sinusoidal function with a global scaling symmetry, and also connected to stu-
dents incorrectly relating a (supposed) lower kinetic energy to a shorter local
wavelength [113, 203]. (b) The “axis shift” phenomenon, similar to students’
representations in [114, 113, 201]. The wave function is depicted on different
levels before and after the barrier.

Some authors have explored the extension of this conception in empirical

studies with undergraduates; for example, in Ref. [114] it is reported for 40%
of n = 105 second-year students and for 21% of n = 80 first year ones in two
different universities in Sweden; in Ref. [123] for 58% of n = 132 engineering
and physics undergraduates of University of Colorado. The authors of Ref.
[201] have also highlighted that this conception appears to be particularly
resistant to being changed through instruction.
Some students develop an inappropriate model of transmission and reflection
at one-dimensional barriers, even in the time independent case when the energy
is fixed, as being caused by particles having a range of possible energies, so that
only the particles with sufficient energy will be transmitted [113, 206]. This
appears to be often suggested by an incorrect use of the uncertainty principle
[123]. According to studies, a significant percentage of undergraduates holds
this mental model; for example, in Ref. [123] it is reported for 27 − 30%
of n = 132 second-year students at the University of Colorado. A variant

28
3.1. Research on student difficulties

of this model is that, for a single particle tunneling, only the “part” of the
particle with sufficient energy will pass the barrier [114, 206]. This is of course
related to a concrete interpretation of the wave function, which is assimilable
to matter-wave conceptions discussed in Sec. 3.1.1.1.

3.1.1.5 Simple one-dimensional bound systems

It has been reported by several authors [204, 205, 208] that many students in-
terpret simple one dimensional binding potentials, such as the infinite square
potential, as concrete bi-dimensional objects, such as holes or wells. According
to Brookes and Etkina [204] this conception is directly evoked by the metaphor-
ical language used in quantum physics (potential wells, potential steps, energy
levels) which transforms, in the minds of students, in a too-powerful analogy.
In this sense, this conception could also be related to the perception of poten-
tial barriers as concrete obstacles, producing friction when penetrated, which
was described in the previous section. In Ref. [208], 36% of students in a
class, and 27% in another, both of second year engineering undergraduates,
when asked what was the most probable point where an electron could be
found in a square potential well, given a certain wave function, displaced the
electron in the vertical dimension.
Students also have problems in interpreting the meaning of wave functions in
confining potentials. It has been reported that they sometimes interpret a
larger amplitude as a higher energy for the particle [114, 187, 208] probably
in analogy with classical waves, and that they do not understand the meaning
of the nodes of the wave function [187]. Styer [188] points out that many stu-
dents believe that energy eigenstates are the only allowed states (instead of the
only stationary states), in analogy to the correct idea that energy eigenvalues
are the only allowed energies, and somehow recreating an historical conception
that existed in the days of “old quantum theory”. In this respect, he sug-
gests devoting more time to at least an initial discussion of time evolution in
introductory courses.

3.1.1.6 Wave functions

Many of students’ difficulties in interpreting wave functions have already been
discussed in the two previous sections. Some general difficulties have also been
reported, which are not related specifically to tunneling or bound systems.
For example, students may have trouble in distinguishing between the wave
function and probability density, and not compute the square of the amplitude
before determining probabilities [187]. Also, according to Refs. [188, 203]
students may form a mental model that the wave function represents a time
average of the positions of particles at different times. The same authors
report that it is relatively common for students to think that a quantum state
is completely specified by its associated probability density, rather than its
wave function.

29
 3. A panorama of research on teaching quantum physics

3.1.2 Difficulties with ideas from classical physics and

mathematical concepts
In this section we will discuss students’ difficulties which are not specifically
related to quantum physics, but which have been associated with significant
obstacles in learning it. Concerning ideas from classical physics, two main
important prerequisites have been highlighted: good functional understanding
of the classical wave model, and of the concept and properties of energy in its
different forms.
Mathematical difficulties are more diffuse and subtle, and several authors have
argued that, with so many conceptual difficulties as we reviewed in the previous
sections, simplifying the mathematics in the high school context becomes a top
priority in order not to present students with impossible tasks [120, 181, 209].
Several studies on students’ mathematical difficulties in quantum mechanics,
such as those by Singh and co-workers [210, 211, 212, 213] refer to specialist
undergraduate or graduate instruction, and to courses held with an approach
based on operators on Hilbert spaces. As such, most of the results contained
in these works are not relevant to the present dissertation. We will refer to
the small number of authors who deal with students’ difficulties in high school
treatment of quantum physics, or in introductory undergraduate courses which
are either algebra-based, or contain only high-school level calculus.

3.1.2.1 Interference and diffraction of waves

Students’ difficulties with waves are clearly relevant in an introductory course
on the subject. In Ref. [187], in a study with n = 38 second year physics
undergraduates enrolled in their first quantum physics course, the correla-
tion of students’ background knowledge of classical waves on their final grades
was higher than the correlation with their math scores. In two related works
[132, 214] McDermott and co-workers investigated in depth difficulties and con-
ceptions of students concerning interference and diffraction phenomena in the
perspective of the introduction of quantum physics. In Ref. [132] they report
that, among n = 46 mixed students, part enrolled in an introductory course
on wave optics, and part in a modern physics course, about 25% believe that,
for the two slits interference experiment, fringes remain identical when one of
the slits is closed; and 20% are convinced that each slit contributes to half the
interference pattern. In the single sit diffraction setup, 20% believe that reduc-
ing the slit width would produce a narrower central maximum, and 25% think
that the diffraction pattern is produced solely by the light which is bended or
scattered from the slit edges. Some students think that the slit width must
be less than the wavelength of light for producing diffraction, while conversely
other students think that, if slit is thinner than the incident wavelength, no
light passes at all and a dark screen results (see also Ref. [22] on this kind of
depictions). The study also highlights that students have difficulties in relat-
ing the origin of maxima and minima to differences in optical path lengths,

30
3.1. Research on student difficulties

and several problems in distinguishing the models of wave and ray optics, and
correctly applying them in their respective contexts. For example, in single
slit diffraction, many imagined the central maximum as due to the shadow of
the screen. Problems related to the simultaneous application of the wave and
ray models are also extensively studied in Ref. [215]. In Ref. [214], 40% of
n = 510 students enrolled in an introductory calculus-based course in wave
optics believed that, in single slit diffraction, the slit width should be less than
the wavelength of light for diffraction to occur; and 30% thought that the dis-
tance between minima would remain the same if half the slit is covered. These
percentages reduced greatly (both to about 10%) after a series of tutorials on
interference and diffraction. In the post test after such tutorial, authors also
asked students questions about the positions of maxima and minima in multi-
ple (more than two) slits interference, obtaining positive results. For treating
this problem, the authors only use a semi-quantitative analysis based on path
length and phase difference, without introducing phasors to represent the am-
plitudes associated to each optical path; they report that reasoning based on
phasors seems too abstract for students, and that, in a few trials that were
performed, many students (about 30%)seem to confuse the angle between two
phasors corresponding to adjacent slits with the spatial angle of deviation of
the optical path from the perpendicular to the screen.

3.1.2.2 Energy concepts

Some authors have argued that the usual teaching approach to the concept of
energy at introductory level in classical physics is inadequate to the task of
favoring a transition to quantum mechanics [216, 217]. In fact, on one hand
energy concepts are usually overshadowed in introductory mechanics by those
of force and acceleration; on the other hand, when energy is introduced, its
crucial properties of being a state property for the system [217, 218]; and of
being always conserved, while changing form or being transferred from one sys-
tem to another, are not sufficiently highlighted. The issues of what is the most
effective approach to the teaching of energy-related concepts, and of providing
students, right from the beginning of instruction, with a consistent picture and
description of energy constitute a very active research field in physics educa-
tion [62, 216, 218, 219, 220, 221].
Some of the relevant problem of students with energy were implicitly high-
lighted in Sections 3.1.1.4 and 3.1.1.5 in the context of conceptual difficulties
with the tunnel effect and the treatment of bound systems. In Ref. [203]
the author highlights that students who correctly interpret the wave function
diagrams in the case of tunneling are consistently associated with having a
mental model of energy as something which is always conserved, while those
displaying the conception of energy loss in tunneling often think of energy as of
a quantity which can vanish in nothingness. Some authors have also reported
that in quantum physics problems students confuse kinetic, total and potential
energy, and may misinterpret the total energy as kinetic [113, 203]. In Ref.

31
 3. A panorama of research on teaching quantum physics

[114] the author proposes to complement, in the discussion of stationary one

dimensional problems such as bound systems and tunneling, the traditional
diagram of potential energy with a graph of the “local kinetic energy” E − V of
the quantum object (which is negative in classically forbidden regions) to help
students distinguish between the forms of energy involved in the problem.

3.1.2.3 Probability and statistics

Students’ models about probability and statistics are clearly relevant in the
learning of introductory quantum theory [187, 206, 208, 222, 223]. In Ref.
[208] the authors studied the conceptions on these subjects of second and
fourth year engineering undergraduates, in the perspective of instruction in
quantum theory. Among n = 18 students of the advanced Physics 420 course
for engineers at University of Maryland, 61% showed the gambler’s fallacy, i.e.
thought that the outcome for flipping an individual coin depends on the results
of previous coin-flipping events; and 27% thought that if a coin were flipped
100 times, results would be exactly 50 heads and 50 tails. They also report
that students may not understand at all the concept of a probability distribu-
tion, in general or as applied to the position of an object. Thus, they suggested
to start with a tutorial on the probabilistic representation of classical system,
for example drawing a probability distribution for the possible positions of a
classical oscillating pendulum, using video analysis tools. However, a possible
drawback of this approach is that it could reinforce the semiclassical concep-
tion of the wave function as an average of particle positions at different times
[115, 188].
In Ref. [187], student difficulties in quantum physics associated to underlying
problems with probability and statistics are thoroughly investigated; the au-
thors report that often students do not understand the concept of a probability
distribution, thinking of it as a “place” or “area”; they do not understand how
it should be normalized, or how the expectation value should be computed;
they confuse expectation value with standard deviation or variance. In the
same study, among n = 38 students taking an introductory quantum course at
Ohio State University, 32 either had mixed ideas or had no clue why we need
to use probability in quantum measurement. Finally, the study also reported
students’ difficulties in applying probability rules for determining the outcomes
of successive events, which are relevant for example in the interpretation of ex-
periments of the Stern-Gerlach type.
The interplay between students’ understanding of probability and quantum
physics is very complex, and possible difficulties are both conceptual, and
specifically related to formalism. Here we limited ourselves to problems that
are likely to appear in introductory high school courses (at least in the Italian
system, students in the final year of science-oriented high schools are expected
to have a reasonable functional understanding of probability distributions and
densities) but several others appear at higher levels of instruction in quantum
physics [187, 188, 210, 222, 223].

32
3.2. Overview of educational perspectives

3.1.2.4 Complex numbers and functions

Complex numbers and their algebraic and geometric properties are a difficult
subject, which historically was not covered in high school courses; this is the
main reason why most elementary introductions to quantum physics do not
explicitly use complex numbers or functions. However, in the last 20 years
curriculum reforms in mathematics have included such topics in secondary in-
struction, so it is possible to start introducing elementary methods of complex
algebra in the physics curriculum. For example, in the sum over paths ap-
proach, complex numbers are typically represented by rotating arrows [4, 5, 6];
but, in some cases, such as in problems including more than one particle, in
which rules for the “multiplication of arrows” have to be given [4]; or, in our
own approach, in the case of the sum over paths representation of tunneling in
classically forbidden regions, it would be very helpful to be able to switch be-
tween the graphical depiction with arrows and the formal representation with
complex numbers.
In Ref. [187] it is reported that even students who were successful in mathemat-
ics introductory courses on complex numbers may have difficulties in applying
these concepts to the context of physics. For example, even if they know the
relationship between the trigonometric and exponential representations of com-
plex numbers, they may not recognize that e−iφ , where φ is a real variable, is an
oscillatory complex function, and may describe it as exponentially decaying. In
individual interviews, some students complain that, although they think they
do understand mathematical concepts such as complex numbers, the language
used in physics to describe and manipulate such concepts is obscure to them.

3.2 Overview of educational perspectives

3.2.1 General approaches to the disciplinary content
In this section we discuss the broad “families” of approaches which are present
in the current scientific debate about the teaching of quantum physics at an
elementary level following in its general lines the classification of Ref. [15]. The
sum over paths approach, and the authors proposing it, will not be considered
here, since they will be extensively reviewed in the next chapter.

3.2.1.1 The historical approach

A great number of researchers in the educational literature, starting perhaps
from Ernst Mach [61, 224] more than a century ago, have claimed that instruc-
tion in physics, and science in general, should adopt an historical approach.
The essential points raised by researchers to support such a claim can be sum-
marized as follows [61]:
1. Introducing an historical perspective in science teaching can link the dis-
ciplinary content to the context in which it was first produced, providing

33
 3. A panorama of research on teaching quantum physics

richer and more meaningful learning patterns, and favoring a deeper un-
derstanding of the disciplinary content;

2. The argument of cognitive recapitulation, that is, the idea that the indi-
vidual growth of knowledge may follow in its general lines the same path
as the original accumulation of collective knowledge. This is perhaps the
main reason why the historical approach is also called genetic [225, 226].

3. Studying the history of physics is revealing of the epistemological nature of

the discipline, of its methodological rules which may remain implicit using
other approaches, and of the process of formation of scientific truth.

In the historical approach to quantum physics, concepts are gradually built

following the development path of the discipline. Historical experiments in-
compatible with a classical conceptual picture are analyzed in sequence, at-
tempting to lead students to retrace the steps of the “scientific revolution”
that brought to the edification of a new theoretical paradigm. This approach
has been traditionally used in the university courses and textbooks of introduc-
tory quantum mechanics [227, 228] although, in this context, it is gradually
being replaced by axiomatic formulations. A more recent book introducing
quantum physics from an entirely historical perspective is Ref. [229]. In the
literature, several positive results were reported in teacher education [230, 231],
and indeed, in our own experience, teachers often ask being instructed using an
historical approach (see also Ref. [1] for the same issue concerning relativity).
In high school textbooks, the historical path is usually simplified, and often
stops before the introduction of the Schroedinger equation. In some cases,
over-simplification may result in giving students a deformed and simplistic vi-
sion of the development of physics, which may also favor misunderstandings of
the disciplinary content [232].
Some of the proposals, which will be discussed in Section 3.2.2, of teaching
quantum physics by revisiting its historical debates and controversies can be
seen as a renovation of the historical approach.

3.2.1.2 Dirac or “spin first” approach

The starting point of the general approach variously called “spin first”, Dirac,
or sometimes “conceptual” [15, 233, 234, 235, 236] is the consideration that the
heart of quantum theory lies in the principle of superposition, applied to a de-
scription of the states of a system based on abstract vectors on Hilbert spaces,
together with the description of physical quantities by operators. According
to the authors of the mentioned papers, such concepts are irrenunciable even
in an elementary introduction to quantum physics, and two-state systems offer
the possibility of introducing them without excessive mathematical difficulties.
An important element in their line of argumentation is that many difficulties
in students’ learning originate from the ambiguity between quantum operators
and the classical concepts of position and momentum, while spin operators are

34
3.2. Overview of educational perspectives

not as easily confused with classical analogues. More in detail, two main vari-
ants can be identified for this approach in current research [237] which should
not be considered as competing, as there have been attempts as a synthesis
[238].

ˆ The proposal advanced by the University of Udine [15, 234, 239, 240], which
originated from the work of Ghirardi and co-workers [241, 242], begins by an-
alyzing the phenomenology of polarization of light, in the context of simple
experiments with polarizers and birefringent materials in which polaroids
play the dual role of preparation and measurement devices. Starting from
simple experimental situations which are realizable in high school laborato-
ries, these authors develop with students a minimal amount of formalism,
using the Dirac representation, and successively use it for interpreting more
advanced experimental setups, such as the Mach-Zehnder experiment with
intermediate polarizers [243]. Also, the formalism can be used as a concep-
tual bridge to introduce the generalization to continuous variables. How-
ever, Michelini and co-workers do not propose to introduce a full treatment
of spin using the mathematical structure of Pauli matrices.

ˆ In the proposal of Pospiech, the phenomenology of spin and its formal treat-
ment using Pauli matrices are introduced from the start [235, 244, 245]. In
this simple, yet full-featured Hilbert space model of quantum theory, a for-
mal basis can be given to several important concepts; for example, the un-
certainty principle can be immediately connected to the non-commutativity
of operators. Experiments using the Stern-Gerlach apparatus or its vari-
ants can help connecting the formal and experimental dimensions. Also,
the discussion can include several concepts which are important in the cur-
rent understanding of quantum physics, such as entangled states, the EPR
paradox and experimental evidence of non locality.

At university level, some textbooks adopting this general perspective are those
by Townsend [233] and Sakurai [246]. In the book Ref. [12], which will also
be discussed in Section 4.3.4 of the next chapter, D. F. Styer starts with the
treatment of spin 1/2 systems to discuss the Stern-Gerlach experiment and
other important results such as nonlocality tests, before switching to the sum
over paths perspective.
Consolidated positive results have been reported in teacher education [247,
248, 249] and encouraging data has been collected in trials with high school
students [239, 250, 251].

3.2.1.3 Proposals based on quantum field theory

Some authors have suggested to start at elementary level with teaching quan-
tum field theory directly, rather than first-quantization non relativistic theory;
the most well known among such proposals are those advanced by Hobson

35
 3. A panorama of research on teaching quantum physics

[252, 253] and, in Italy, by Giliberti [254, 255]. Recently, important contribu-
tions on the possibility of providing an educational reconstruction of quantum
field theory have also been given by Bertozzi and other authors of the Univer-
sity of Bologna [116, 256, 257]. Both the proposals of Hobson and Giliberti
are characterized by suggesting a precise ontology for quantum theory, one in
which quantum objects (both massless, such as photons, and massive, such
as electrons) are identified as discrete excitations in a field. In particular, in
the approach of Giliberti a special unifying role is attributed to the Klein-
Gordon equation, in a perspective which in some respect echoes Feynman’s
construction of scalar quantum electrodynamics in the third and fourth chap-
ter of QED ([4], see Section 4.2.2). Giliberti’s approach is the only one, in
the international panorama, which attempts a full educational reconstruction
of quantum physics, including derivation of the Schrödinger equation, and
conceptual issues such as wave-particle duality, the uncertainty principle, and
discussion of atomic structure, starting from QFT concepts [116]. One of the
basic guiding principles in this approach is that history should not necessar-
ily be the sole reference guide in education: a general coherent framework of
quantum physics must be presented to students, independently of how it is
grown. Authors proposing this view argue that since quantum field theory
is the best understanding of nature we currently have, and is not equivalent
to first quantization quantum mechanics, all possible efforts should be made
to teach it the former, rather than the latter, to students. One of the ad-
vantages which have been reported for this approach is that QFT is naturally
formulated in four-dimensional space-time [252], and so can bypass some of the
counterintuitive features of non relativistic quantum mechanics, which requires
a 3N-dimensional configuration space. Again, encouraging results have been
reported for this approach with student teachers [258].
Recently, the proposal of Hobson has been criticized by the author of Ref.
[259] (who expresses similar objections to those contained in Ref. [116]) on
the grounds that no ontological content can be attributed to QFT, nor it is
necessary to do so. Hobson, however, responded [259] defending the consis-
tency of a field ontology. The idea of attributing a special ontological status to
the Klein-Gordon equation, which is a central idea in Giliberti’s approach, is
also examined in depth, and globally rejected, although with some cautionary
remarks, in Refs. [116, 256].
Some other authors suggest strategies for introducing quantum field theory
concepts in high school, for example concentrating on Feynman diagrams [260],
and processes and symmetries in particle physics [261, 262], but they do not
elaborate a full educational reconstruction of quantum theory, and the con-
ceptual structure upon which their proposals are based remains unspecified.
More details on some of these works can be found in Ref. [116].

36
3.2. Overview of educational perspectives

3.2.1.4 Formal-analogical approaches

In this approach [263, 264, 265] the point of departure is an in-depth study of
some classical system which may be seen as formally analogous to a quantum
one. Typically, oscillating strings or membranes; or linearly coupled oscillator
systems are considered; and the analogy is drawn between normal modes and
quantum eigenstates. The very straightforward treatment of linearity and su-
perposition is the strongest point in favor of this approach. Experimentations
in the instruction of future teachers have been performed using this approach,
obtaining good results particularly with teachers with a degree in mathematic
[264].
In order not to leave students with the false impression of a complete continu-
ity in the transition from classical to quantum mechanics, this approach must
be complemented with an in-depth and detailed analysis of the epistemological
crisis of classical mechanics in the early XX century. The “Electronium” pro-
posal by Niedderer and co-workers [168, 169, 170, 165, 171] which we discussed
in Section 3.1.1.2 can be considered a form of analogical approach.

3.2.2 Epistemological perspectives

3.2.2.1 Epistemological and linguistic renovation

The most important attempt at renovating the epistemological content of quan-

tum physics after its definitive consolidation was started by Bunge [146, 266]
in the late 1960’s, and prosecuted by Lévy-Lebond [148, 149, 150, 267]. a cen-
tral element in Bunge’s attempt at a new axiomatization of quantum physics
which could get rid of wave particle duality, and also of subjective elements
contained in the Copenhagen interpretation [144] was the concept of “quan-
ton”. The quanton is neither a particle nor a wave or field, but a different
kind of object, associated with a certain distribution of the variables of po-
sition and momentum. Bunge’s attempt to lay the foundations of quantum
physics has been criticized [268], but, regardless of whether his proposal is
considered successful, his ideas had a long lasting influence in physics educa-
tion. Starting from Bunge’s proposal, Lévy-Leblond has argued in favor of a
profound linguistic renovation in formulating and teaching quantum physics,
eliminating or substituting terms reminiscent of Bohr’s own view on the re-
lationship between quantum physics as language [149] such as “wave particle
duality” and “observable”. Together with Balibar he produced an introduc-
tory university-level textbook on quantum mechanics constructed according
to these principles [267], and he progressively refined them in successive sci-
entific works. For example, he suggested replacing “observables” with “proper-
ties”; “non-separability” with “implexity”; “non locality” with “pantopy” [149]
and so on. Also, Lévy-Leblond provided a further specification of the content
of the term “quanton”. According to him, classical objects have either the
characteristics of being localized in space-time, and discrete in count number

37
 3. A panorama of research on teaching quantum physics

(particles); or being continuously extended in space-time, and formed of an

inseparable continuous (waves). Quantum objects or quantons, on the other
hand, are discrete in count number, but continuously extended in space-time
[150, 144]. Within his educational works, meaningful analogies can be found,
representing how the concept of “wave-particle duality” should be overcome
and historicized: the analogy of the cylinder interpreted as a rectangle-circle
[267] and of the platypus described as a duck-mole [269].
Bunge and Lévy-Leblond’s proposal of epistemological and linguistic renova-
tion has been adopted, entirely or in part, by some authors in physics educa-
tion [75, 24, 257], and elements of it are beginning to appear in high school
textbooks, starting from France [144]. In our own view, Lévy-Leblond’s spec-
ification of the meaning of the “quanton” concept summarized above applies
perfectly to the quantum object as described by Feynman’s formulation1 . In
this respect, we found Lévy-Leblond’s analogies mentioned above, and in par-
ticular the second one, very effective in educational practice.

3.2.2.2 Teaching through historical-epistemological debates

In recent years, a number of authors in science and physics education have
given rise to a substantial movement of innovation of historical approaches to
teaching. According to these researchers, it is not sufficient to consider the
“genetic” approach as a strategy, more or less effective, for teaching science
content, but the focus should be shifted on realizing that knowledge about
science (about the ways, and contexts, in which concepts and theories were de-
veloped), is as important as knowledge in science (about the concepts, norms,
laws comprised in a discipline) [270]. The necessary knowledge for teachers
to promote student’s learning about a discipline, the evolution of its discourse
and the dynamics of scientific revolutions (i.e. promoting “discipline-culture”,
[271]) has been called Cultural Content Knowledge (CCK) [271, 272] as com-
plementary to Pedagogical Content Knowledge (PCK) [273]. An important
aspect of CCK is the ability of teachers to provide an adequate representation
of the dynamics through which different theories replaced one another in his-
tory. In fact, it is often remarked [271, 274, 275] that textbooks, even those
following an historical approach, often provide a fictitiously linearized account
of such historical development, presenting the introduction of new theories as
the sole logical possibility, and minimizing or neglecting the role of dissidents,
and controversies.
According to these principles, some authors [75, 191, 274, 276, 277] have sug-
gested involving students in the discussion of historical debates and controver-
sies which played an important role in the early stages of development of the
theory, both in the perspective of conceptual change [191] and with the aim of
challenging an authoritative and exclusive image of science, in which a unique
1
In our work we consistently used the term “quantum object”, but not “quanton”, partly
because of not wanting a too close identification with Bunge’s epistemological views, but in
part simply because the Italian term quantone sounds somewhat awkward.

38
3.2. Overview of educational perspectives

point of view is legitimate and possible [75]. Some of the controversies that
have been proposed for discussion in educational setting are those between
Bohr and Heisenberg on the meaning of the uncertainty principle and the
gamma ray microscope [75, 191, 277]; between Bohr and Einstein on realism
and the fundamental nature of quantum theory [75, 274, 277]; between Millikan
and Ehrenhaft with respect to the determination of the elementary electrical
charge [276]; or the philosophical debate about the reduction of chemistry to
physics [274].

3.2.2.3 Taking or not taking a foundational position

A fact that appears more evident in quantum theory than in other areas of
physics and science, is that its foundational bases are not settled: foundations
of quantum physics are still a very active research field; there is no consensus
between physicists on what should be the definitive (if ever there will be one)
interpretation of the theory [278, 279, 280], nor there is consensus between
researchers in education on the attitude that teachers should take towards
foundational matters [114, 281]. In a series of detailed studies on this mat-
ter, Baily and Finkelstein [111, 125, 282] do not reach the conclusion that the
teaching of a particular interpretation of quantum theory should be preferred;
however, they warn that teachers should be aware of the possible consequences
of their own foundational attitudes on students’ learning. For example, they
examine two apparently very similar slides on the one-dimensional problem
of the infinite square potential well, one of which is produced by a teacher
with an agnostic foundational attitude, who does not explicitly discuss aspects
related to interpretation during the course. The slide, they argue, contains
subtle hints encouraging a realist perspective that in the end can produce a
measurable influence on the persistence in students of classical deterministic
conceptions. Cheong and Song [143] suggest that educators take a suspensive
attitude towards foundational matters, which is not to be confused with an
agnostic stance. According to them, the suspensive perspective consists in
carefully interpreting crucial experiments relevant to wave particle duality and
entanglement, separating and clarifying what is the noncontroversial part of
quantum theory, and revealing aspects which are controversial, and debated
between alternative interpretations of the theory.
Other authors have suggested specific foundational attitudes. Often the Copen-
hagen interpretation is taken as a primary reference [114] although it is not al-
ways clear what physicists call the “Copenhagen interpretation” precisely refers
to [283]. Starting instruction using the statistical interpretation [284] has also
been suggested [23, 137]. Some other examples and suggestions [24, 144, 145]
have been discussed in the previous sections.

39
 3. A panorama of research on teaching quantum physics

3.2.3 Complementary strategies and tools

Independently of their global approach, many authors, as we have done, have
introduced in their educational proposals the discussion of modern experi-
ments, used computer simulation or online resources such as virtual labora-
tories. Here we schematically present some reference coordinates for these
complementary educational strategies and tools.

3.2.3.1 Discussion of modern experiments

Many authors have suggested use modern experiments (to define “modern” in
this context, we consider experiments which have been first performed after
the 1950’s), often from quantum optics, in the context of elementary intro-
ductions, to provide students with additional insight on the conceptual and
foundational aspects of quantum physics. A number of suggestions involved
the use of the single photon Mach-Zehnder interferometer, both in the polar-
izing and non polarizing versions [117, 137, 243, 285, 286, 287, 288]; or the
Merli-Missiroli-Pozzi (or Tonomura) two slit experiment with one electron at
a time [131, 133, 189, 252, 254, 289]. Others included the Grangier experi-
ment on photon indivisibility [290], experimental violations of Bell inequalities
and their relations with the EPR paradox and nonlocality [12, 244, 288, 291],
the Zhou-Wang-Mandel experiment [292], experiments with interference and
diffractions with other massive quantum objects than electrons [254] and sev-
eral others.
In Refs. [293, 294], which we only discovered at the end of this dissertation
work, the authors advance a proposal for the introduction of the photon con-
cept in an undergraduate introductory course, by performing and analyzing of
some modern experiments, which has several points of contact with our own:
they use the Grangier experiment [295] for establishing the existence of the
photon, the single-photon Mach-Zehnder (both polarizing and non polarizing)
to demonstrate the photon interfering with itself, and the Hong-Ou-Mandel
setup [296] for showing interference between identical photons. For the analy-
sis of these experiments, they use a matrix approach and the Dirac formalism.

3.2.3.2 The use of simulations

The role of interactive simulations has been studied in depth by educational
research, and the general conclusion has been drawn that such instruments,
especially when used in the context of guided exploration and inquiry activi-
ties, can offer significant support to the learning process, and provide relevant
educational gains [297, 298, 299, 300, 301]. This is particularly true for sub-
jects, such as quantum physics, in which real experimental activities in the
laboratory are not easy to realize, at least for advanced problems.
The number of researchers who focused on developing computer simulations of
quantum phenomena for educational use has been steadily growing in the last
20 years. Several online repositories of educational simulations for quantum

40
3.2. Overview of educational perspectives

mechanics exist. To cite only the most famous ones, QUVIS of University of
St. Andrews [288, 302, 303, 304]; Physlets of the comPADRE group [305, 306];
PhET of University of Colorado [139, 178, 181, 182]; Visual Quantum Me-
chanics from Kansas State University [135, 176, 183, 184]. In 2014 an inter-
national group of physicists, affiliated with Multimedia Physics for Teaching
and Learning (MPTL) performed a survey of simulation repositories and other
multimedia-based learning resources for quantum physics, producing a detailed
review of the best materials available [307], to which we refer the interested
reader. Also, several other authors among those cited in the previous sections
[137, 167, 206, 286], and many of those who will be reviewed in the next chap-
ter [5, 12, 151, 308], which deal specifically with proposals based on the sum
over paths approach, have used simulations in their educational approaches.

3.2.3.3 Remote and “realistic virtual” laboratories

Online laboratories allowing users to visualize real apparatuses and obtain re-
alistic experimental results are emerging in recent years as a valuable resource
for teachers, alternative or complementary to both real experiments and simu-
lations, or virtual simulated laboratories. “True” remote laboratories are those
in which real, functioning and operational apparatuses are synchronously con-
trolled through the visual interface, and results are displayed precisely as they
were obtained in the experimental run that the user has initiated. On the
other hand, using our own nomenclature, as the field is very young and still
lacks an university accepted terminology, we call “realistic virtual laborato-
ries” those in which the user is visualizing real apparatuses which actually
exist (or have existed) in some research structure, but the visualized results
are recorded ones, corresponding to the parameters set by the user. True re-
mote laboratories are certainly more fascinating for students, but also have
the non insignificant disadvantage that only one user at a time can perform an
experimental run, and the risk for teachers is to find the apparatus “busy” at
the time scheduled for the activity with students. Remote laboratories avail-
able online are still few, but the list is quickly growing. Among those which
are relevant for the teaching of quantum physics we mention the one of the
University of Munich containing, among others, apparatuses for interference
and diffraction of electrons, and for the photoelectric effect [309]; iSES of the
University of Prague, allowing to perform the Franck-Hertz experiment as well
as experiments on the diffraction of light from micro-objects [310]; and iLabs
[311], a project of MIT which is more centered on electronics, but also allows
to perform experiments on Bragg diffraction of electrons and radioactivity. A
very good realistic virtual laboratory, which we used in our sequence both in
high school and with student teachers is QuantumLab of The University of
Erlangen [312], which offers several quantum optics experiments, such as the
Grangier and Hong-Ou-Mandel ones.

41
3. A panorama of research on teaching quantum physics

42
Chapter 4
The sum over paths approach

4.1 Fundamentals of Feynman’s path integral

formulation
As stated by Schulman in the first line of his famous book “Techniques and
Applications of Path Integration” [313], the best place to find out about path
integrals in in Feynman’s original 1948 paper [2]. In this article, Feynman
presents a new formulation of quantum physics, proving its equivalence with
the ordinary description in terms of the Schrödinger equation. In this section
we will briefly review the basics of the path integral formulation, following the
general lines of reasoning of Feynman’s original paper.

4.1.1 Classical and quantum probabilities

Feynman starts his argumentation with a reflection about the use of probabil-
ity in classical and quantum physics. In order to make such comparison, we
have to make the assumption that classical experiments, for example detection
experiments, can have probabilistic results. This is of course reasonable, be-
cause of epistemic (i.e. due to ignorance) uncertainty (for example, ignorance
of the initial conditions) in classical physics.
Having accepted this, suppose our aim is to compute the probability that, in
an ideal experiment, an event E, which can happen in N alternative (mutually
exclusive) ways b1 , b2 , b3 ...bn starting from the same initial state A, is observed.
Also suppose all the conditional probabilities that the event E happens through
the “channels” bi , i.e. P (E|bi ) are known.
In this case, according to classical physics, a simple rule for computing the
conditional probability P (E|A) holds:

N
X
P (E|A) = P (E|bi )P (bi |A) (4.1)
i=1

43
 4. The sum over paths approach

The essential difference between classical and quantum physics,as stated by

Feynman, is that in the former Eq. 4.1 is always true, while in the latter it
is ordinarily false. Indeed, assuming that the channels bi are indistinguishable
in the given experimental setup (see also Section 4.9), in quantum physics the
conditional probabilities can be expressed as

P (E|bi ) = |ϕ(E|bi )|2 P (bi |A) = |ϕ(bi |A)|2 (4.2)

Where ϕ(E|bi ) is a complex number called the probability amplitude that event
E happens through the channel bi . The correct quantum rule for computing
the conditional probability of event E starting from the initial state A is then
 2
XN 
P (E|A) =  ϕ(E|bi )ϕ(bi |A) (4.3)
 
 
i=1

That is, the total probability that an event E, which can happen through N
alternative channels b1 , ...bN , is observed, is given by the square modulus of the
sum of the complex probability amplitudes ϕ(E|b1 )ϕ(b1 |A), ..., ϕ(E|bn )ϕ(bN |A).

4.1.2 Intuitive approach to the path integral formula-

tion
Several authors (see for example Refs. [3, 314]) present an intuitive introduc-
tion to the path integral formulation based on a generalization of Young’s two
slit interference experiment. Consider a setup in which a source at xS emits
approximately mono-energetic quantum objects, for example electrons.1 The
apparatus also includes an intermediate screen B with two slits b1 and b2 , and a
detector on a final screen at point xD . If the two slits are imagined as pointlike,
the probability of the event of a detection at xD is given by the contribution
of two alternative possibilities: that the electron passes through the slit b1 or
through the slit b2 . Accordingly, Eq. 4.3 applies:
 2
XN 
P (xD |xS ) =  ϕ(xD |bi )ϕ(bi |xS ) (4.4)
 
 
i=1

Suppose now to make the experiment more complicated, adding more slits
b1 , b2 ...bN to the screen B, and also one more screen C with slits c1 , c2 ...cN as
in Fig. 4.1. Now by repeatedly applying E. 4.3 the conditional probability
P (xD |xS ) can be written as
 " #2
X X 
P (xD |xS ) =  ϕ(xD |cj ) ϕ(cj bi )ϕ(bi |xS )  (4.5)
 
 
j i

1
The choice of a source emitting quantum objects with well defined energy allows us to
ignore the time description in the intuitive introduction [314].

44
4.1. Fundamentals of Feynman’s path integral formulation

Figure 4.1: The multiple interference setup with successive slits which is used
as a “gedankenexperiment” in the intuitive introduction to the path integral
formulation

Eq. 4.5 can be expanded to take the form of a sum of amplitudes from all
different paths leading from xS to xD passing through any combination of two
successive slits, i.e.

X
P (xD |xS ) = | ϕpath (xS , bi , cj , xD )|2
i,j

with ϕpath (xS , bi , cj , xD ) = ϕ(xD |cj )ϕ(cj bi )ϕ(bi |xS ) (4.6)

The argument can immediately be generalized to a number of screens between

the source and the detector, and a number N of slits on each screen, which both
approach infinity; in this case, screens and slits can be thought as fictitious
devices, simply representing a discretization of space. If the index µ counts the
number of paths γµ (x) leading from xS to xD , the probability can be expressed
as  2
X 
P (xD |xS ) =  ϕ(xS , γµ (x), xD ) (4.7)
 
 
µ

4.1.3 Postulates of the path integral formulation

In a more precise description of the Feynman formulation, time must also be
considered. However, in order to keep things simple, we will limit ourselves
to one spatial dimension. We define P (xS , tS , xD , tD ) as the probability that
a quantum object initially at space-time point (xS , tS ) (which does not nec-
essarily represent a “source”, however we will keep the notation for clarity) is
detected at space-time point (xD , tD ). We can write the first postulate of the
path integral formulation of quantum mechanics as: [2, 315]

Postulate 1. The probability that a particle initially at (xS , tS ) is detected at

(xD , tD ) is
P (xS , tS , xD , tD ) = |K(xS , tS , xD , tD )|2 (4.8)

45
 4. The sum over paths approach

where the amplitude K(xS , tS , xD , tD ) can be expressed as the sum of all the
single-path amplitudes, one for each possible continuous and piecewise differ-
entiable [316] path leading from (xS , tS ) to (xD , tD ), or
X
K(xS , tS , xD , tD ) = K(xµ (t)) (4.9)
all µ paths

and the xµ (t) are all the possible paths such that xµ (tS ) = xS , xµ (tD ) = xD .

Where K(xS , tS , xD , tD ) is usually called the propagator from (xS , tS ) to

(xD , tD ). The central point is of course how K(xµ (t)) is computed, i.e. how each
path contributes to the total amplitude. Starting from the assumption that
Postulate 1 is satisfied, and that the classical trajectories correctly describe
the motion of a particle in the formal limit h → 0, it is possible to guess the
second postulate [314]:

Postulate 2. All paths contribute to the propagator with complex numbers of

equal amplitude, but whose phase is the classical action in units of the quantum
of action ~:
i
K(xµ (t)) = Ce ~ S[xµ (t)] (4.10)
where C is a normalization constant, and the classical action S is defined as
the time integral of the Lagrangian:
Z tD
S[xµ (t)] = L(x, ẋ, t)dt (4.11)
tS

where L(x, ẋ, t) is the Lagrangian of the system. In order to provide a

workable definition of the propagator, the usual procedure is to discretize the
time interval (tS , tD ) in N intervals of length τ = ti+1 − ti , to integrate at each
time instant ti over all possible positions x(ti ) of the particle, and finally to
take the limit as τ → 0:
Z Z Z
i P dx(t1 ) dx(t2 ) dx(tN −1 )
K(xS , tS , xD , tD ) = lim ... e ~ i S(x(ti ),x(ti+1 )) ...
τ →0 A(τ ) A(τ ) A(τ )
(4.12)
where A(τ ) is a normalization constant which is needed for convergence, which
will be computed in Sec. 4.1.6. Eq. 4.12 is essentially the definition of the
time propagator, which is usually written in more compact form as
Z xD
i
K(xS , tS , xD , tD ) = e ~ S[x(t)] D[x(t)] (4.13)
xS

The problem of defining, in general, a suitable Lebesgue measure on the space

of all paths raises, in general, severe mathematical difficulties that go well
beyond the scope of this dissertation [316].

46
4.1. Fundamentals of Feynman’s path integral formulation

4.1.4 From the propagator to the wave function

From Eq. 4.13 and the additivity of the classical action S computed over
successive time intervals, which is evident from its definition in Eq. 4.11,
a composition rule can be derived for combining the amplitudes of events
occurring in a temporal succession:
Z
K(xS , tS , xD , tD ) = K(xS , tS , x(tc ), tc )K(x(tc ), tc , xD , tD )dx for tS < tc < tD
(4.14)
Since the propagator K(xS , tS , x(tc ), tc ) for given (xS , tS ) and fixed time tc is
just a complex-valued function of the space coordinate x, we may define such
function as the wave function ψ(x, tc ) of the quantum object at time tc and
rewrite Eq. 4.14 as
Z
ψ(xD , tD ) = ψ(x, tc )K(x(tc ), tc , xD , tD )dx for tc < tD (4.15)

The meaning of Eq. 4.15 is that, if the set of complex amplitudes for a quantum
particle is known for all points of space at a given time instant tc , then the past
(t < tc ) no longer counts: the wave function ψ(x, tc ) contains all that is needed
to predict future probabilities, and the question of how the wave function itself
has come into place becomes irrelevant to the future evolution of the system.
This can also be seen as a formulation of Huygens’ principle, which, as noted
by Feynman [2] is exactly valid for nonrelativistic quantum particles, while it
is only approximately valid in optics.
It may be interesting to write down explicitly the propagator for a free particle.
The derivation can be found in Ref. [3] and basically consists in substituting
the free particle Lagrangian L = 21 mẋ2 in Eq. 4.12 and solving a set of gaussian
integrals.

− 12
im(xD − xS )2

2iπ~(tD − tS )
K(xS , tS , xD , tD ) = exp (4.16)
m 2~(tD − tS )

It may be noted that the square modulus of the time propagator is a Gaussian
function in space, whose variance increases as the time interval tD −tS becomes
larger. This fact has relevant consequences, even from an educational point of
view. First of all, it means that a quantum object which is initially perfectly
localized in space (i.e. whose wave function may be thought a Dirac Delta
δ(xS , tS )) will evolve in time to produce a Gaussian probability density; and
secondarily, that a free particle probability density which is initially Gaussian
will remain so under time evolution, only spreading more and more as time
advances. This explains the prevalence of wave functions which are initially
“Gaussian wave packets” in the time-dependent treatment of non relativistic
quantum systems [317].

47
 4. The sum over paths approach

4.1.5 The correspondence principle

In the classical limit, that is when the action S of the system becomes very
large with respect to the quantum of action ~, the Feynman formulation has
to reproduce the results of classical mechanics. In other words, the corre-
spondence principle must hold: in order to have a consistent picture of the
world, classical physics must somehow turn out to be a limit theory of quan-
tum physics.
This can be seen in the following way: consider two neighboring paths x(t)
and x0 (t), where x0 (t) = x(t) + η(t) where η(t) is a function having small values
for all times in (tS , tD ). Then we can expand the action S[x0 (t)] in powers of
η(t) as
Z
0 δS[x(t)]
S[x (t)] = S[x(t) + η(t)] = S[x(t)] + η(t) dt + O(η 2 ) (4.17)
δx(t)
So that the phase difference between the two paths is approximately
Z
1 0 1 δS[x(t)]
∆φ = (S[x (t)] − S[x(t)]) ' η(t) dt (4.18)
~ ~ δx(t)
In the formal limit ~ → 0, the phase difference between neighboring paths is
large even though the parameter η(t) is small; thus destructive interference
occurs. The only exception is the classical path xCL (t), for which the action is
stationary, i.e. δS[x(t)]/δx(t) = 0 and no cancellation with neighboring paths
occurs.
We can describe the argument in a more pictorial way: consider representing
each contribution to the propagator as a vector in phase space. Each vector
has the same length, but a different phase angle, given by φ = S[x(t)] ~
. The
propagator is proportional to the vector sum of these contributions. Now,
consider paths x and x0 close to the classical path xCL . Since S[xCL ] is an
extremum of S, S[x] and S[x0 ] will not differ appreciably from each other, or
from S[xCL ]. As a result, the phase of each of these paths will be similar, the
vectors will be nearly parallel and the corresponding vectors will therefore add
in phase, contributing significantly to the total propagator.
Now, suppose that the two paths x and x0 are far away from the classical
path. We are now distant from the extremum of S, and so S[x] is likely to
be significantly different from S[x0 ]. The phase of these contributions will
therefore be very different, and the vectors are likely to almost cancel each
other, producing a very small contribution to the propagator. A graphical
description of the situation is presented in Fig. 4.2.

4.1.6 Equivalence to the Schrödinger equation

A crucial step in the presentation of Feynman’s approach is to demonstrate
its equivalence to the usual description of quantum physics in terms of the
Schrödinger equation. We here follow closely Feynman and Hibbs’ proof [3].

48
4.1. Fundamentals of Feynman’s path integral formulation

Figure 4.2: The“Cornu spiral” representing a typical situation for the sum of
amplitudes over paths, in this case for a free particle. The classical path xCL is
at a minimum of the action and is almost in phase with the neighboring paths

We start by defining the propagator for a small time τ . From Eq. 4.12 we can
write such object as

  
1 iτ x + x(t + τ ) x(t + τ ) − x τ
K(x, t, x(t + τ ), t + τ ) ' exp L , ,t +
A ~ 2 τ 2
(4.19)
By using Eq. 4.15 we can write evolution for small times for the wave function
as
Z ∞
ψ(x, t + τ ) = K(x, t, y, t + τ )ψ(y, t)dy (4.20)
−∞

Next we explicitly write the Lagrangian in the form L = mẋ2 /2 − V (x, t) and
make the substitution y = x + η so that the integral over y becomes an integral
over η.

∞
mη 2
Z  
1 iτ  η τ
ψ(x, t + τ ) = exp − V x + ,t + ψ(x + η, t)dη (4.21)
−∞ A ~ 2τ 2 2 2

Since reasonably the propagator will not displace the particle too much in a
very small time τ , we expect that only an interval [−η0 , η0 ], for some value of
η0 which depends on τ , will contribute significantly
q to the integral in in Eq.
4.21. More precisely, if η gets larger than about τm~ , the phase of the first
part of the exponential inside the integral varies rapidly. Therefore, most of
the contribution to the integral will be given by values η 2 ≤ τm~ . This means
that when (in the following passage) we will expand the expression to first
order in τ , we will also need to keep terms up to second order in η. By doing

49
 4. The sum over paths approach

so we obtain:

∂
ψ(x, t) + τ ψ(x, t) =
∂t 
1 ∞ imη2 η2 ∂ 2
Z  
iτ ∂
e 2~τ 1 − V (x, t) ψ(x, t) + η ψ(x, t) + ψ(x, t) dη
A −∞ ~ ∂x 2 ∂x2
(4.22)

Where we have used the substitution τ V x + η2 , t + τ2 ' τ V (x, t) which is

correct to first order in τ . By equating the terms of zeroth order in τ in Eq.

4.22 e obtain the value of the normalization constant A
r
2iπ~τ
A= (4.23)
m
which is used in the derivation of many propagators, such as the free particle
one in Eq. 4.16. To proceed further we compute two Gaussian integrals:
Z ∞
imη 2
ηe 2~τ dη = 0 (4.24a)
−∞

 m  12 Z ∞ imη 2 i~τ
η 2 e 2~τ dη = (4.24b)
2iπ~τ −∞ m
so that we can rewrite Eq. 4.22 as

∂ iτ ~ ∂2
ψ(x, t) + τ ψ(x, t) = ψ(x, t) + ψ(x, t)V (x, t) − ψ(x, t) (4.25)
∂t ~ 2im ∂x2
Eq. 4.25 is equivalent to

~2 ∂ 2
 
∂
i~ ψ(x, t) = − + V (x, t) ψ(x, t) (4.26)
∂t 2m ∂x2

which is the time dependent Schrödinger equation.

4.2 Sum over paths for photons: Feynman’s

QED
In 1983 Richard Feynman gave a series of lectures on quantum electrodynamics
(QED) for laymen at the University of California at Los Angeles in memory
of his friend Alix G. Mautner. These lectures were then collected in a book,
“QED: the strange theory of light and matter” [4] which is to this day one of
the most successful divulgation books on physics of all times. The aim and
attitude of the author in preparing the lectures are well described by the fol-
lowing long quote:
“That’s my position: I’m going to explain to you what the physicists are doing

50
4.2. Sum over paths for photons: Feynman’s QED

when they are predicting how nature will behave, but I’m not going to teach
you any tricks so you can do it efficiently. You will discover that in order to
make any reasonable prediction with this new scheme of quantum electrody-
namics, you would have to make an awful lot of little arrows on a piece of
paper. It takes seven years - four undergraduate and three graduate - to train
our physics student to do that in a tricky, efficient way. That’s where we are
skipping seven years of education in physics: by explaining quantum electro-
dynamics to you in terms of what we are really doing, I hope you will be able
to understand it better than do some of the students!”
What Feynman is saying here is that he will try to convey the conceptual con-
tent of quantum electrodynamics, without delving into mathematical details
and techniques, but also without modifying significantly the content itself in
the attempt of making it accessible to a larger audience, a goal that can rightly
be called the holy grail of physics divulgation and education for non special-
ists. Whether the attempt can be considered a complete success is debatable,
but certainly Feynman’s book has been an invaluable source of inspiration for
much educational research, including Taylor’s and our own. Some additional
considerations are needed on the method presented in QED since, as shown
in Sec. 4.1.6, Feynman’s original path integral machinery can be seen as an
alternative way of finding solution to the time dependent Schrödinger equation
which, being non relativistic, certainly cannot be expected to describe photons.
I will divide the discussion into two different subsections because, although the
whole of Feynman book can be considered a consistent educational reconstruc-
tion of quantum field theory, the material presented in chapters 1 and 2 can
be understood in much simpler terms than the content of chapters 3 and 4.

4.2.1 QED chapters 1 and 2: sum over paths for the

monochromatic photon
In the first two chapters of QED, electrons never appear (except in the descrip-
tion of single photon detectors, i.e. photomultipliers) and the author presents
a sum over paths model to deal with the quantum behaviour of light in the
familiar (for a physicist at least) cases of transmission, reflection and refrac-
tion from a dielectric medium, interference and diffraction, mirror reflection.
In particular, to introduce the non classical features of the photon, Feynman
starts from the motivating example of partial reflection of light from glass layers
of slightly different width. He starts from considering the “thinnest possible”
layer of glass, whose width is much smaller than the photon wavelength, and
continues the mental experiment by increasing the width of the glass sheet of
a very thin amount, of the order of one quarter wavelength of the photon. Just
about any classical theory that could be devised to deal with a particle-like
photon would predict the number of transmitted photons to decrease as the
width of the glass is increased; what actually happens instead is that the num-
ber of transmitted photons oscillates with the sheet width, going from zero to

51
 4. The sum over paths approach

the double of an average value. Another point that certainly deserves to be

highlighted in Feynman’s book is the coherent presentation of the uncertainty
principle using the diffraction of light form a slit of variable width, seen from
the sum over paths perspective (pp.54-56).
The model introduced in QED makes the explicit simplifying assumption of
ignoring polarization; also, in these chapters only strictly monochromatic pho-
tons are considered. Feynman proceeds by first stating few, very simple and
general rules, and later adding more specialized prescriptions when required
by the discussion of the examples presented. Summarizing, we can gather the
following set of rules from the first two chapters of QED:

1. Complex amplitudes for individual events are represented as vectors, and

the probability that an event happens is equal to the square of the length
of the corresponding vector.

2. If an event can happen in alternative ways, the individual vectors from

the possible alternatives must be summed, finding a resultant vector (the
“final amplitude”) which must then be squared to find the probability of
the event.

3. In particular, the probability of detecting at a point xD a photon emitted

from a source xS is calculated by adding up all the individual amplitude
vectors resulting from all the possible paths from the source to the detector,
and taking the square of the length of the resultant vector.

4. For each of the individual possible paths of the photon, the direction of the
amplitude vector rotates as the hand of an imaginary stopwatch, making
one full turn each time the photon advances in its paths of one wavelength
(in a particular medium).

5. The modulus of each individual amplitude vector is inversely proportional

to the length of the path. This is required for the conservation of flux (pp.
41,74).

6. If the photon in a certain path is reflected from a mirror, the corresponding

vector inverts its direction (takes a π additional phase) after the reflection.
If reflection from a dielectric material happens, the π additional phase is
computed for “external” reflection (from a lower to a higher refraction in-
dex), but not in the opposite case of “internal” reflection (p.27).

7. The modulus of each individual amplitude vector also depends on trans-

mission and reflection coefficients from layers of dielectric materials encoun-
tered (pp.65-72).

8. If an event happens in two successive “phases” (for example, the photon

can reach an intermediate slit in several ways, and after the slit can reach
a detector in several ways again) the complete amplitude can be computed

52
4.2. Sum over paths for photons: Feynman’s QED

by multiplying the amplitudes in the sense of complex numbers, i.e. taking

the product of their moduli and adding up their phases(p. 60).

While the first three rules are very general, and essentially contain the basic
principles of the sum over paths approach presented in Sections 4.1.1 and 4.1.2;
and rule 8 is a formulation of the Huygens principle, some of the remaining
rules (and most notably rules 6 and 7) do not have an exact counterpart
in the path integral formalism. To understand the theoretical bases of the
picture presented in the first two chapters of QED, we start by reminding that
the simplified spin zero photon thereby presented is a massless Klein-Gordon
particle, described by the equation

1 ∂2
∇2 ψ(r) − ψ(r) = 0 (4.27)
c2 ∂t2
which is of course also the wave equation of scalar optics. The fundamental so-
lutions to this equation for a pointlike source (i.e. in spherical coordinates with
angular symmetry) are spherical waves. Looking for monochromatic (definite
frequency) solutions leads to the Helmholtz equation

∇2 ψ(r) + k 2 ψ(r) = 0 (4.28)

This means that a basic quantum theory of the photon (ignoring polariza-
tion) can simply be constructed by taking the wave equation of scalar optics
[318] and interpreting the wave function ψ as an ordinary quantum ampli-
tude. There are actually some subtleties in this operation (see for example
Ref. [319]) due to the fact that the resulting probability density in Eq. 4.27 is
not, in general, positive definite, and requires adopting the so called “Feynman-
Stueckelberg” interpretation of negative energy solutions as particles propagat-
ing backwards in time. These complications, however, are irrelevant to the case
of definite frequency solutions.
The Green function for the Helmholtz equation in three dimensions is [320, 321]

eik|(r2 −r1 )|
G(r1 , r2 ) = (4.29)
4π|r2 − r1 |

Where we have chosen the outgoing wave. Eq. 4.29 is sometimes also called
the fixed energy photon propagator in coordinate space [322]. The method of
obtaining the amplitude ψ(rD ) for a photon emitted from a certain source at
rS by summing the contributions of Eq. 4.29 from all possible paths from rS
to rD may be called “stationary path integral” [323].
Since the amplitude associated to the photon is essentially the same thing as
the value of the complex “disturbance” [318] at a given point in space of the
classical wave, it is no surprise that the former inherits all the known rules
valid for the latter: phase shift upon reflection, transmission and reflection
coefficients.

53
 4. The sum over paths approach

Feynman’s chapters 1 and 2 of QED is suggestive of a very promising strat-

egy for introducing quantum physics at an elementary level: to start with the
quantum nature of light, and to start with the time independent case. In
this way, it is possible to revisit the whole familiar phenomenology of wave
optics, essentially using the same mathematical methods which are already
known, but switching the language used to the quantum perspective: talk
about photons rather than waves, compute probabilities of detection instead
of intensities, add up elementary vectors over all possible paths, rather than
considering phase differences between interfering waves.
As a side note, we observe that a well known analogy exists between the homo-
geneous Helmholtz equation Eq. 4.28 and the time independent Schrödinger
equation for a constant potential:

2m(E − V )
∇2 ψ(r) + ψ(r) = 0 (4.30)
~2
Such correspondence is a defining property of the field of matter-optics
[324, 325]. In section 4.4 we will show how such analogy can be exploited to
construct a sum over paths method for the elementary treatment of confined
systems in the time independent case.

4.2.2 QED chapters 3 and 4: Scalar electrodynamics

and the Klein-Gordon equation
Most of the material covered in the second part of QED is perhaps less rel-
evant from an educational point of view, although it still contains important
suggestions. Among these, a discussion of “which way” experiments from the
sum over paths point of view, which we will expand in Sec. 4.9 as it consti-
tutes a central element in our own approach; and a discussion of the different
behaviour of bosons and fermions under exchange of two particles, and the
Pauli exclusion principle, which Feynman treats (pp. 110-113) momentarily
abandoning his simplifying assumption of considering electrons as spin zero
particles. In section 5.3.7 we will fully exploit this educational strategy, con-
necting it to experimental results which were not yet available when QED was
published.
Most of the second part of the book is devoted to photon-electron interactions,
and it may be argued that the climax of the book is reached (pp. 103-110) when
the initial, motivating example of partial reflection of photons from a layer of
glass of variable width is discussed from a fully microscopic, time-dependent
point of view, including absorption and re-emission af photons from electrons.
Feynman starts his discussion by presenting three basic “actions” combining
which all the phenomena involving photon-electron interactions can be repre-
sented:
1. A photon goes from place to place
2. An electron goes from place to place

54
4.2. Sum over paths for photons: Feynman’s QED

3. An electron emits or absorbs a photon

After that, he provides rules for drawing space-time Feynman diagrams, and
computing amplitudes for lines and vertexes. A very important concept in-
troduced here, that can actually be used in educational practice, is that the
total amplitude for a diagram involving more than one particle is obtained by
multiplying the amplitudes corresponding to individual quantum objects (see
Section 5.3.7).
Since the treatment here is fully time-dependent, the underlying mathemati-
cal machinery is a proper path integral formulation, similar in principle to the
basic one summarized in Section 4.1, although of course it is the relativistic
version. The theory which Feynman is presenting is scalar QED [326] (since the
electron spin is neglected), but due to the simplification of ignoring the photon
polarization also, the relevant free fields can be described as two Klein-Gordon
particles, one of zero mass (the photon) and one of finite mass (the electron)
[327]. Feynman’s presentation in these two chapters is appropriate for a wide
audience, but the concepts introduced are very advanced, and the theoretical
justification for some of the statements provided is not always transparent; this
has motivated the author of Ref. [327] to write a technical paper explaining
to specialists what exactly Feynman is explaining to non specialists in some
particular passages of QED. We discuss only one example, precisely Figure 56
on page 90 of QED, which we reproduce along with its original caption as Fig.
4.3.

Figure 4.3: Reproduction of Fig. 56 on p. 90 of QED. The original caption

says: When light goes at speed c, the “interval” I equals zero, and there is
a large contribution in the 12 o’clock direction. When I is greater than zero,
there is a small contribution in the three o’clock direction inversely proportional
to I; when I is less than zero, there is a similar contribution in the nine o’clock
direction. Thus light has an amplitude to go faster or slower than speed c, but
these amplitudes cancel out over long distances.

In the figure reported Feynman explains that the photon has non zero
amplitudes of following space-time paths in which its speed is lower or greater
than the speed of light, but these two kinds of amplitudes cancel out with

55
 4. The sum over paths approach

one another. We follow closely the explanation in Ref. [327], omitting some
mathematical details.
With the conventions used in QED, the propagator for a Klein-Gordon particle
to change its space-time position by an interval x = (c4t, 4r) = (x0 , x) is
(in the next two equations, until Eq. 4.33, we set c = 1 for uniformity with
the literature on quantum field theory).
d4 p ip·x
Z
1
G(x) = 4
e (4.31)
(2π) p − m2 + i
2

Referring specifically to the case of the photon, we set m = 0. Then E = |p|

and Eq. 4.31 becomes
d3 p −ipx e−i|p||x0 |
Z
i
Gγ (x) = − e (4.32)
2 (2π)3 |p|
For the case x0 6= 0, this integral can be evaluated as
i 1 i 1
Gγ (x) = − 2
= − (4.33)
(2π)2 x2 − x0 (2π)2 (4r)2 − (c4t)2
Thus, if (4r)2 > (c4t)2 (i.e. v > c) then the amplitude is GF ∝ −i; while if
(4r)2 < (c4t)2 (i.e. v < c) then GF ∝ +i.
These amplitudes correspond to the two little arrows pointing to the right and
to the left in Fig. 4.3.
The remaining case to consider is (4r)2 = (c4t)2 , that is v = c. In this case
x20 = x2 and the photon propagator can be written as
d3 p −ipx e−i|p||x|
Z
i
Gγ (x) = − e . (4.34)
2 (2π)3 |p|
By passing into spherical coordinates Eq. 4.34 can be transformed into (see
Ref. [327] for details)
Z ∞
i 1
Gγ(x) = − du sin(u)(cos(u) − i sin(u)) (4.35)
(2π)2 |x|2 0
with u = p|x|. The integral in Eq. 4.35 is the sum of two parts, the first of
which is bounded, while the second one approaches infinity. Thus the ampli-
tude in this case diverges, and corresponds to the big arrow pointing upwards
in Fig. 4.3.

4.3 Educational proposals based on the sum

over paths approach
4.3.1 The Taylor et al. approach
From 1996 to 2005, E. F. Taylor taught an undergraduate course on quantum
physics at MIT called “Demistifying quantum mechanics” in which the subject

56
4.3. Educational proposals based on the sum over paths approach

was presented from the point of view of Feynman’s sum over paths approach.
In 1998 he and coauthors Vokos, O’Meara and Thornber published in Com-
puters in Physics an article describing the development of the course and the
materials used [5] which had a profound impact in the international physics
educational research community. At the time, Taylor was already well known
for his and J. A. Wheeler’s revolutionary approach to the teaching of special
relativity [49]. In the following years, Taylor co-authored with J. Ogborn an
article introducing the same perspective in high school education [7], worked on
the connection between Feynman’s approach and the principle of least action
[328] and made the materials for his introductory course in quantum physics,
including the educational simulations, freely available on his website [329].
Taylor et al. first paper [5] is a detailed account of the undergraduate course at
MIT, which started from Feynman’s presentation of the quantum behaviour of
light (chapters 1 and 2 of QED) and proceeded with a full educational recon-
struction of the path integral approach, integrated with student activities on
educational simulations. More in detail, the course is divided in the following
steps:

A. The photon. The authors consider Feynman’s example of partial reflection

of light from a glass sheet of variable width, and use it to introduce the
basic rules of the sum over paths approach. A computer simulation is used
to show students the sum of arrows corresponding to different paths, and
the formation of the Cornu spiral.

B. The electron. The analogy between the behaviour of photons and elec-
trons is noted, and the rotation rate for the “quantum stopwatch” in the
case of the electron is guessed from the classical principle of least action.
The treatment is, from the start, time dependent; the action computed
along each path is expressed by Eq. 4.11.

C. The wavefunction. The concept of wave function is introduced by apply-

ing, with the help of a computer simulation, the sum-over-paths algorithm
to a particle at two sequential times. The wave function is defined as the
collection of arrows that represent the electron at various points in space
at a given time.

D. The propagator. Here students derive, using heuristic reasoning and with
the help of simulations, the form of the propagator for the free particle. This
part of the sequence is quite complicated for several reasons. First of all,
computing the time propagator essentially corresponds to the procedure of
taking the continuum limit for a multiple integral with a discrete set of
integration variables described in Eq. 4.12. The normalization constant
A which appears in such equation, and was derived as Eq. 4.23, can only
approximately be guessed through qualitative reasonings. Other proper-
ties that the aithors use with students for the heuristic derivation, such as
that the propagator must preserve the shape of a distribution of amplitudes

57
 4. The sum over paths approach

which is constant over all space, can only be justified as postulates. Finally,
there are also problems related to the numerical computation, because sam-
pling only a finite number of paths produces in this case misleading results.

E. Propagation in time of a nonuniform wavefunction. Having ob-

tained the propagator for the free particle, students use it to evolve in time
non uniform wave functions. This gives them an intuitive feeling for the
quantum time evolution, which can be used for introducing them to the
Schrödinger equation.

F. Wavefunctions in a potential. Except for the uniform potential, in

all other cases (e.g. the infinite square and harmonic wells) the relevant
propagator is built into the computer program that the authors developed
for use with students. In this step, students use the software to explore the
time development of an initial wavefunction in a potential.

G. Bound states and stationary states. Using the same software tool
as in the previous step, students are led to discover, by trial and error,
stationary states for a given potential. In this case, in fact, time evolution
only produces an overall rotation in unisons for all the arrows of the initial
wave function. Also, discrete energy levels can be identified as the different
possible speeds which are allowed for such unison rotation.

The structure of the course presented by the authors is very modern, alternat-
ing traditional lectures, worksheets for guided inquiry activities with educa-
tional simulations, open questions and discussions. Unfortunately, the original
executable files have been developed within the cT programming environment
[330] now largely obsolete, and having never been updated or ported to new
systems have suffered technological aging and are nowadays basically unusable.
At the end of the paper, the authors report the positive reactions of students
to the course, and enumerate the advantages of adopting the sum over paths
approach. Among these, certainly are worth mentioning the conceptual sim-
plicity of the method; the possibility of building the quantum description of
nature as an extension of the classical one, and to switch back and forth at
any time between the former and the latter; the possibility to provide a self-
contained, conceptual, non-mathematical introduction to the subject for those
who do not need to use quantum mechanics professionally.
Also, they enumerate a few of the perceived disadvantages of the new approach.
In particular, the most relevant difficulty they report is with the treatment of
confined systems, citing two reasons: that the propagators for most confining
potentials are not analytically known, and that the approach adopted requires
to delay the discussion of bound states to the end of the course. Such difficulty
of the traditional sum over paths approach, with which we definitely agree, is
the main reason that motivated us to take a different route in the treatment
of bound states for confined systems, which will be discussed in Section 4.4.

58
4.3. Educational proposals based on the sum over paths approach

To this, we may also add that in Taylor et al. original course there is practi-
cally no attempt at making students actually solve exercises and find answers
to quantum physics problems: almost all activities revolve around the manip-
ulation of simulations. Thus one may question whether, besides conceptual
understanding, whose importance certainly cannot understated, students are
actually able to compute anything with pen and paper in quantum physics
after having attended the course. To this question, also, we tried to provide a
positive answer in the design and testing of our own proposal (see Chapter 5).
In Ref. [7], Ogborn and Taylor present a version of the educational path more
tailored to a public of secondary school educators. In this work they strengthen
the connection between the sum over paths approach and the classical prin-
ciples of least action and least time, proposing to teach, right from the start,
Newton’s law of motion as a consequence of the classical principle of least
action. In this paper they also discuss the intuitive assumption that, using
the formalism of path integrals for massive particles, the amplitude vector cor-
responding to a particular space time path does approximately one complete
turn for each De Broglie wavelength. Despite its apparent evidence, the state-
ment is not so obvious and must be proven (see also Ref. [325]) which they do
in the Appendix.

4.3.2 Ogborn, Dobson and the Advancing Physics AS

project
In year 2000 the British Institute of Physics inaugurate the Advancing Physics
project [11], an advanced course in physics for high schools, consisting of two
concatenated texts (called AS, or “advanced subsidiary”, and A2) designed to
attract students to physics, and to give them a good basis for their future
progression in the subject at university level [331]. One of the chief editors of
the project was J. Ogborn, who shared many of Taylor’s views on the impor-
tance of the principle of least action in physics education [7, 332, 333] and took
the route of the sum over paths approach for quantum physics in “Advancing
Physics AS”. At the time of its launch, in the academic year 2000/2001, it was
distributed to about 350 schools in Britain. The inauguration of the project
was also accompanied by a paper by A. Dobson in the Physics Education
journal [6] in which he presented the main innovative points of the project
concerning the teaching of quantum physics.
Being a comprehensive high school physics course, “Advancing Physics AS” is
able to make a drastic choice aimed at rendering the introduction of quan-
tum physics with the sum over paths approach easier and more natural: to
treat the entire chapter on the wave behaviour of light, before the introduction
of the photon, using the method of phasors. Indeed, wave optics and quan-
tum physics are grouped in a single super-chapter called “Waves and Quantum
Behaviour”. Phasors are actually an extremely convenient tool for treating
interference and diffraction phenomena, and are used for performing compu-

59
 4. The sum over paths approach

tations in other textbooks, including the hugely successful Halliday-Resnick-

Walker [195]. However, “Advancing Physics AS” is arguably the only textbook
in which the choice is motivated by the successive introduction of the photon
using the sum over paths language developed by Feynman in the first two
chapters of QED. In this case, in performing the transition to a new language
described in Section 4.2.1, it is possible to retain exactly the same mathemat-
ical tools used for the discussion of the wave phenomenology of light [153].
As it is to be expected for a large scale educational experiment, “Advancing
Physics” offers a much wider range of materials to choose from with respect
to other proposals. Suggested experimental activities for the quantum physics
section include single photon detection, reflection gratings for selective light
diffraction, measurement of the h costant using LEDs of different colors, and
electron diffraction. Educational simulations have been updated since 2000,
and are realized within the ModellusTM environment [334] which is well sup-
ported. The course also includes a few exercises, such as the calculation by
hand, using contributions from a sample of all possible paths, of the amplitude
for the reflection of a photon from a mirror surface. The mathematics used
in the course is not complicated, but enough “to make it physics and not just
some kind of fairy story” [6].
The aforementioned article by Dobson discussing the course has several merits.
First of all, it makes clear that although talking about photons, the instructor
should not emphasize the corpuscular aspect, and possibly not even refer to
photons using the “particle” term. In the section titled “Not wave behaviour,
not particle behaviour, but quantum behaviour” (which is also the title of one
of the Section of Advancing Physics AS) he evidently echoes the ideas of Levy-
Leblond [148, 150] in stressing the importance of using new concepts, and even
new words, for describing quantum entities. Another important section of Ref.
[6] is the one titled “Frequently Asked Questions” in which the author reports
on the recurring doubts and difficulties of students, from pilot tests performed
in 1999, before the beginning of the large scale adoption of the course in high
schools. All the questions reported are relevant to any educational approach
using the sum over paths perspective, but especially two of them were thor-
oughly considered when designing our own teaching-learning path. The ques-
tion “What do the photons really do?”, concerns the possibility that students
interpret the prescription of exploring all paths as just an abstract mathemat-
ical construction, masking the underlying fact that photons actually take one
path and one only. Avoiding the emergence of this conception was the factors
which initially motivated us to include in design of our sequence the discussion
and simulation of the Mach-Zehnder experiment with individual photons and
removable arms (see Section 5.2.5). The question “Do the photons all travel at
different speeds?”, can be raised by students when using the Feynman model
to solve optical problems, since the photons arrive to the detector traveling
along paths of different length. Answering this question requires a deep reflec-
tion on the concepts of coherence and monochromaticity and, as we will see in

60
4.3. Educational proposals based on the sum over paths approach

more detail in Section 5.2.4 also calls into question the uncertainty principle.
Expanding slightly on the answer suggested by Dobson et al., one may recall
that the assumption of definite frequency implies, in quantum mechanics, an
infinite uncertainty on the emission time. It follows that the interfering paths
do not represent the photon traveling at different speeds, but rather having
been emitted at different times (to within its coherence time, which, in the
case of a perfectly monochromatic object, would be infinite).
In our case, in order to avoid sources of confusion with the concept of time in
an essentially time-independent approach, we took the stance of consistently
using path length, rather than travel time, as the correct quantity to be used
to compute the phase of the elementary amplitude (see Section 5.2.4) although
the two quantities are of course proportional and are often used interchange-
ably in the literature.
Still alive 15 years after its inauguration, and used in 2006 by around 25%
of British high schools including an upper level physics course, “Advancing
Physics AS” is certainly the largest scale experimentation of the sum over
paths method in secondary education that has ever been performed. However,
it seems that its results, and specifically those concerning the quantum physics
section, have been only superficially evaluated. In 2003 a first survey of the
satisfaction of students and teachers involved in the project was performed
[335]. Overall approval from teachers was at 85%. Concerning students, sat-
isfaction with the examination and its outcomes was very high, with about
90% of responses being positive; and approval of the textbook and course ma-
terials ranked at 80%. The report, however, did not mention specifically the
quantum physics part of the course. In 2006 Ogborn produced a document
[13] providing a very qualitative and subjective evaluation of the results of
the quantum physics part of the project, based on his experience of discussion
with teachers and students at meetings and on mailing lists. In this report,
the quantum physics course is described as invariably stimulating for students
and teachers, who engage in animated and participated discussions. According
to the author, teachers appear to be satisfied of the approach proposed, and
of the methods and materials used. However, Ogborn also reports a tendency
from both teachers and students to give to the ideas of quantum physics an
overconcrete interpretation. For example, they often associate a traveling pha-
sor to a traveling photon, rather than associating phasors to paths, or produce
concrete models of the phasors as rotating wheels.

4.3.3 Rinaudo and the the proposal of Torino

In Italy, the sum over paths approach to quantum physics is traditionally [15]
tied to the Unversity of Torino and the physics education group of G. Rinaudo.
Building on Feynman’s QED [4], on Taylor’s proposal [5] and also on the work
of Fabri, who had developed a similar approach for high school education in
1996 [337]. Rinaudo’s group designed and developed its own proposal, testing
it both with pre-service teachers and directly with high school students, start-

61
 4. The sum over paths approach

ing from 2000 [9, 308, 338].

A central theme in this proposal is a radical criticism to the traditional teaching
of the concept of energy in high school. According to Rinaudo, the excessive
emphasis on the concept of force as the primary explanatory principle of dy-
namics obscures the value of energy as an universal state variable for physical
systems, and as an unifying concept for all the branches of physics. A similar
criticism applies to the teaching of momentum, which is usually taught as an
ancillary concept of velocity, while is discovered in quantum physics to be the
true fundamental dynamical property of objects in motion. Consequently, the
introductory parts of the proposal of the Torino group are centered around the
role of energy in quantum physics, its relation to frequency, and the Planck
relation E = hν.
Other interesting characteristics of the proposal of Torino are the rigor in
the enunciation of principles, for example in precisely formulating the distinc-
tion between allowed (and thus possible) and forbidden paths [308], or the
distinction between objects of the physical theory and abstract mathemati-
cal constructs (such as the phasor arrow). The urge of adopting a precise
and unambiguous language leads Rinaudo and co-workers to employ the term
“quantum object”, rather than wave or particle, for speaking of electrons and
photons, as suggested by Levy-Leblond [148, 150]. Also, the authors often
highlight the possibility of providing students with a non misleading visualiza-
tion of the model of quantum theory as one of the most important advantages
of the sum over paths approach, a theme which is also central in our own pro-
posal.
The proposal includes several suggested experimental activities, such as the
qualitative observation of Wien’s law, diffraction from a slit of variable width,
refraction from a dielectric material and experiments with lenses and diffrac-
tion gratings. The authors have also produced several simulations of relevant
examples solved with the sum over paths method. The simulations are very
detailed and contain a large amount ov visualized data and several adjustable
parameters; they are realized with MS ExcelTM , so their predictable lifetime
should be almost infinite; on the other hand, they are visually not very attrac-
tive.

4.3.4 Other proposals

In this section, we briefly account for other proposals based on the sum over
paths approach which have been advanced by the physics education commu-
nity.

ˆ In 2000, J. Hanc of the Technical University of Kosice in Slovakia, who also

collaborated with Ogborn and Taylor [333], started developing his version of
a course for both university and high school students based on Feynman’s
approach [8, 336]. The content of the course is similar to “Demistifying
quantum mechanics”, but in the university version includes the treatment

62
4.4. Treatment of time independent problems

of more advanced subjects such as the Noether theorem, Feynman diagrams

and particle-antiparticle creation and annihilation. Simulations are of good
quality and visually engaging, programmed in Java. In Ref. [14] the author
tries to deal with the difficulty of teaching energy levels and eigenstates
from using Feynman approach highlighted by Taylor ([5], see Section 4.3.1)
by developing a sum over paths-like procedure for solving by trial and error
the discretized version of the time independent Schrödinger equation.

ˆ From 1989 to 2009 D. F. Styer of Oberlin College, Ohio, taught a course

of modern physics for non physicists titled “The Strange World of Quan-
tum Mechanics”. Lectures of the course were condensed in a book with the
same title [12]. The book starts presenting experimental results such as the
Stern-Gerlach apparatus with repeated measurements, treats the Einstein-
Podolsky-Rosen paradox, some of its variations and the experimental test
of nonlocality. About halfway it introduces the sum over paths formalism,
and proceeds using the Feynman approach to treat interference experiments,
wave particle duality, as well as modern applications such as quantum com-
puting, cryptography and delayed choice experiments. Styer is also the
author of Ref. [327] discussed in Section 4.2.2.

ˆ In Argentina, M. de los Ángeles Fanaro, M. R. Otero, and M. Arlego have

been working for several years on experimentations based on the sum over
paths approach [10, 289, 151]. Contrarily to many other works, their studies
do not concern global proposals for the teaching of quantum physics, but
usually concentrate on one or two problems of great significance, adopting
the sum over paths perspective as reference conceptual structure, and ana-
lyze in detail the results of tests, and the process of knowledge construction
in the classroom and in individual discussion groups. For example, in Ref.
[289] they build a sequence of thirteen lessons based on the free evolution
and the two slit experiment only. For these basic situations, they eviscerate
every possible aspect: using photons, electrons or marbles, studying the
classical limit and so on. They use both ModellusTM and GeoGebra [339]
simulations, in which students have the possibility of modifying a great num-
ber parameters and conditions; and they present detailed results on both
students’ conceptions and their emotional responses.

4.4 Treatment of time independent problems

As discussed in Section 4.3.1, a typical difficulty in teaching sum over paths
quantum theory is that the treatment of elementary one-dimensional stationary
problems is not entirely straightforward [5, 336]; for example, finding the eigen-
functions for confining potentials usually requires computing the time depen-
dent propagator, and then determining (often with the help of some software
tool) those initial distribution of amplitudes which, for the given propagator,

63
 4. The sum over paths approach

are stationary in time [5]. Furthermore, this approach forces educators to re-
verse the typical teaching-learning path to quantum physics, introducing the
time dynamics of quantum objects first, and treating stationary systems later.
In our own educational approach [19, 152, 340] we decided to take a different
route. In the spirit of what Feynman did in the first two chapters of QED [4],
where he started by developing a time independent sum-over-paths description
of the monochromatic photon, and only later introduced time evolution, we
initially treat massive quantum objects in one dimensional potentials using a
sum over paths approach at fixed energy, independent of time. Indeed, this is
more than an analogy, because of the formal identity between the Helmholtz
equation and the time independent Schroedinger equation in a constant po-
tential [324]. In this way the conceptual structure of Feynman’s formulation is
entirely preserved, and we are able to treat many confined systems of interest,
as well as the important case of tunneling, without (or before) introducing the
time dependent version, i.e. the sum over paths representation of the time
dependent propagator.
The following section can be considered as the first part of a tentative answer
to research question Q2, i.e. an attempt at recapitulating and condensing in a
compact and general form what is known about the sum over paths approach
to the energy dependent Green function, so that it can be used as a reference
point for designing teaching approaches in educational practice.

4.5 Sum over paths methods for the Green

function
The educational advantage of Feynman’s approach lies mainly in its conceptual
structure. The conceptual understanding of single particle quantum physics
which is provided by the sum over paths approach can be summarized as
follows:

ˆ The quantum object goes through all possible paths from an initial space-
time point (xi , ti ) to a final one (xf , tf ).
i
ˆ A complex number e ~ S[x(t)] (often represented, in educational practice,
by a conventional rotating vector[4, 5]) is associated to each one of the
Rpaths;
tf
its phase angle is proportional to the classical action S[x(t)] =
ti
L (x, ẋ, t) dt calculated along the path.

ˆ The (normalized) sum of all contributes from the possible paths starting
at (xi , ti ) and ending at (xf , tf ) gives the time dependent propagator
K(xi , xf , ti , tf ), which can be understood as the probability amplitude of
finding at (xf , tf ) a quantum object which was initially at (xi , ti ).

ˆ The probability P(xf , tf ) of detecting the quantum object at (xf , tf )

is then computed by taking the square modulus of the propagator, i.e.

64
4.5. Sum over paths methods for the Green function

P(xf ) = |K(xi , xf , ti , tf )|2 .

For time-independent problems, the role of a transition probability amplitude
from point xi to point xf in space, for a fixed energy E, is instead played by
the energy-dependent propagator, or Green function, G (xi , xf , E). The sum
over paths approach at fixed energy, for time independent problems, can be
compactly described as follows:
ˆ The quantum object goes through all possible paths at fixed energy E
from an initial point in space xi (the source) to a final one xf (the
detector).
1
ˆ A complex number, proportional to e ~ S[x] is associated to each one of the
paths; its
R xfphase angle is proportional to the classical abbreviated action
S[x] = xi p(x)dx calculated along the path.
ˆ The sum of all contributes from the possible paths at fixed energy start-
ing at xi and ending at xf gives the energy dependent propagator, or
Green function G(xi , xf , E), which can be understood as the probability
amplitude of finding at xf , independently of arrival time, a particle with
defined energy whose source is at xi .
ˆ The probability P(xf ) of detecting the quantum object at xf is then
proportional to the square modulus of the Green function i.e. P(xf ) ∝
|G(xi , xf , E)|2 . For bound systems, the probability is nonvanishing only
when the energy E corresponds to an allowed energy level.
In this way the conceptual structure of Feynman’s formulation is entirely pre-
served with two main modifications: 1) the action S[x(t)] is substituted with
the abbreviated action S[x]; and 2) all paths connecting xi and xf , indepen-
dently of travel time, are considered, which allows to immediately introduce
the idea, of educational value in itself, that when energy is fixed, time must be
completely unknown.
The Green function, or energy dependent propagator, G(xi , xf , E), defined as
the Fourier transform of the time-dependent propagator, contains all relevant
information for a quantum system. It can be understood as the probability
amplitude for a particle whose source is at xi and whose energy is fixed at
the value E to be detected at xf . For a general one-dimensional system, the
spectral representation of the Green function is
ψk (xf )ψk∗ (xi ) X ψn (xf )ψ ∗ (xi )
Z
n
G(xi , xf , E) = dk + (4.36)
E − Ek + i n
E − E n + i
Where the sum has to be carried out in the discrete part of the spectrum, and
the integration in the continuous part. For systems with a discrete spectrum,
eigenvalues correspond to the poles of the wave function, while eigenfunctions
can be obtained by evaluating the limit
ψk (xf )ψk∗ (xi ) = lim (E − En ) G(xi , xf , E) (4.37)
E→En

65
 4. The sum over paths approach

i.e. are the residues of G(xi , xf , E) at the poles.

4.5.1 The Van Vleck - Gutzwiller propagator

For one dimensional confining potentials, the Green function can be evaluated,
at least in an approximate way, by summing complex amplitudes G(xi , xf , E),
which we will call “elementary propagators” or “elementary amplitudes” over
all classical paths at fixed energy E connecting xi and xf .
Since the energy is fixed, paths of any duration have to be considered, and so
all paths performing an arbitrary number of periodic orbits must be included in
the sum. Often G(xi , xf , E) is defined as the Van Vleck-Gutzwiller semiclassical
propagator [313, 341, 342]:
m 1 π
GV V G (xi , xf , E) = 2 p ei( ~ S(xi ,xf ,E)−µ 2 ) (4.38)
i~ ki kf
where the index µ counts the number of turning points of the classical trajec-
tory which, in this case, is used as a path; and S(xi , xf , E) is the abbreviated
action, in which we have highlighted the dependence from energy, rather than
time. Gutzwiller’s approach allows to prove the generalized WKB formula for
the allowed energy levels of U-shaped potentials:
1
S0 (En ) = 2(nπ + φ) (4.39)
~
Where S0 (En ) is the action for one complete classical orbit inside the potential
at energy En , and φ corresponds to the phase loss at each turning point.
Eq. 4.38 implies φ = π2 , which only produces the exact spectrum for some
potentials, such as the harmonic oscillator and Morse ones; in other cases the
correct phase factor for a given potential (which may depend on the energy
En ) can be assigned a posteriori, after comparing the results obtained with the
exact ones [348].

4.5.2 Sum over paths representation of the Green func-

tion for piecewise constant potentials
For the free particle, the exact energy dependent propagator (the Fourier trans-
form of the free time-dependent propagator) is known to be
m
G(xi , xf , E) = 2 eik|xf −xi | (4.40)
i~ k
This has suggested [19, 343, 344] the following method for piecewise con-
stant potentials in one dimension:
A. Consider all possible paths going from xi to xf , passing through both clas-
sically allowed and classically forbidden zones, with the sole limitation that
a path can reverse its direction only at a point in which the potential is
discontinuous.

66
4.6. One dimensional bound systems in piecewise constant potentials

B. For each path, compute the elementary propagator

m 1
G(xi , xf , E) = 2 ei( ~ S(xi ,xf ,E))
Y
Cν (4.41)
i~ ki ν
P
where the abbreviated action S(xi , xf , E) is computed as the sum j kj ∆xj
over all the zones traversed by the path, the index ν counts the total number
of transmissions and reflections for the path considered; and the Cν are the
usual transmission and reflection coefficients [343]:
k1 − k2 2k1
r12 = t12 = (4.42)
k1 + k2 k1 + k2
:
C. To obtain the green function G(xi , xf , E), add up all the elementary prop-
agators.
The transmission and reflection coefficients Eq. 4.42 are usually obtained
by imposing continuity and differentiability of the wave function for a step po-
tential. In order do so, it is not necessary to introduce the Schrödinger equa-
tion. The coefficients can be derived starting from the elementary propagator
Eq. 4.41 with unknown Cν , and deriving them by imposing continuity and dif-
ferentiability of the resulting Green function for the step potential. However,
in educational practice the coefficients can be derived from an analogy with
optics [324], or simply presented as a postulate.
Also note that our basic propagator Eq. 4.41 only contains the initial wavenum-
1
ber, with the factor ki−1 rather than (ki kf )− 2 as the Van Vleck-Gutzwiller
propagator and the one used in Ref. [344]. This choice allows us to keep
the traditional transmission and reflection coefficients of Eq. 4.42 rather than
modifying them, and also has interesting consequences for the self consistency
of the approach which will be discussed in Section 4.8. In the following we
provide several examples of how the method we have introduced is used; in
particular we discuss one dimensional bound (Sec. 4.6) and open (Sec. 4.7)
systems, obtaining exact results as long as the potentials considered are piece-
wise constant. In Section 4.8 we use an approximation scheme to extend the
method to smooth confining potentials, such as the harmonic oscillator, ar-
riving essentially to the WKB results, although through an unconventional
route.

4.6 One dimensional bound systems in piece-

wise constant potentials
4.6.1 Infinite square well
The system considered is the usual “particle in a box”, a quantum object
confined in a hole of width L with infinitely high potential walls. The solution

67
 4. The sum over paths approach

strategy presented here is rather well known (see for example Ref. [313]) but
we find it useful to briefly recapitulate it to introduce to the language and
notations of a more general approach.
For a fixed energy E, the particle can reach the detector xf starting from a
source at xi through one of four families of paths, depicted schematically in
Figure 4.4. The Green function thus takes the form

Figure 4.4: Principal paths for each family in the infinite square potential well,
and their associated single path energy dependent propagators. The paths
depicted are also valid for the inner region of the finite square well in section
4.6.2.

G(xi , xf , E) =
∞
m  ik|xf −xi | 2 ik(2L−|xf −xi |) ik|xi +xf | ik(2L−|xf +xi |)

 X
= 2 e +r e +r e +e r2p e2ipkL =
i~ k p=0

m e f i + r2 eik(2L−|xf −xi |) + r eik|xi +xf | + eik(2L−|xf +xi |)

 ik|x −x |

= 2 (4.43)
i~ k 1 − r2 e2ikL
In this case, since outside the well the wavenumber approaches imaginary in-
k−iκ
finity, the only reflection coefficient is r = limκ→∞ k+iκ = e−iπ . Notice that
the choice of the minus sign (a “phase loss”) is not arbitrary since as κ → ∞, r
approaches the value −1 from the lower complex half-plane. By direct inspec-
tion of Eq. 4.43 it is apparent that the poles of the Green function correspond
to those values of k for which
2kL − 2π = 2nπ n = 0, 1, 2, .. (4.44)

68
4.6. One dimensional bound systems in piecewise constant potentials

which implies kn = (n+1)π

L
. This allows to immediately obtain the spectrum
of energy levels. Expressed in words, the condition in Eq. 4.44 selects those
values of k for which two paths differing by a basic periodic orbit are in phase.
In order to obtain the eigenfunctions it is sufficient to compute the Green
function G(x, x, E) by evaluating the limit of eq. 4.43 as xf → xi :

m i2kL i2kx i2k(L−x)

1
G(x, x, E) = 2
1 + e − e − e (4.45)
i~ k 1 − e2ikL
The square modulus of bound states wave functions is obtained by taking the
limit in Eq. 4.37

i2kL i2kx i2k(L−x)

1 1 + e − e − e
|ψn2 (x)| = lim k 2 − kn2


=
k→kn 2ik 1 − e2ikL
2 sin2 (kn x) (k 2 − kn2 ) 2
= lim 2ikL
= sin2 (kn x) (4.46)
ikn k→kn 1 − e L

Eigenfunctions are obtained with the correct normalization constant, and can
be chosen to be real (For all one dimensional confining potentials, bound state
wave functions can, and usually are, arbitrarily chosen to be real, see e.g. Ref.
[345]):
r
2
ψn (x) = sin(kn x) (4.47)
L
In Section 5.4.7, Fig. 5.16 we show an example of the GeoGebra simulations
that we use in order to illustrate this case to students.

4.6.2 Finite square well

Now we turn to the case in which the quantum object is within a potential
well of width L, but whose walls have a finite height V , i.e. a potential of the
form (
V if x < 0 ∨ x > L
V (x) = (4.48)
0 if 0 ≤ x ≤ L
In this case a conceptual separation has to be made between the cases that the
position of the source xi and the detector xf are either inside or outside the
well.
For both xf and xi lying inside the well, the Green function G(xi , xf , E) has
formally the same expression as Eq. 4.43, but the reflection coefficient r has
now the form
k − iκ 1 − iγ
r= = = e−2i arctan γ (4.49)
k + iκ 1 + iγ
√ p
with k = ~1 2mE, κ = ~1 2m(V − E). and γ = κk . By making use of the
exponential form of the reflection coefficient in Eq. 4.49 we derive G(x, x, E)

69
 4. The sum over paths approach

as
m  i2(kL−2 arctan γ) −2i arctan γ i2kx i2k(L−x)


G(x, x, E) = 1 + e + e e + e ·
i~2 k
1
· 2i(kL−2 arctan γ)
(4.50)
1−e
The quantization condition can thus be identified as

kL − 2 arctan γ = nπ (4.51)
q
And using the trigonometric identity arctan γ = π2 − arcsin VE we obtain a
rather elegant and concise proof of the quantization condition first derived by
Landau and Lifshitz [320, 346]
r
kL E π
+ arcsin = (n + 1) (4.52)
2 V 2
As it should be, Eq. 4.44 is consistent with Eq. 4.52 if the infinite square well
of Sec.4.6.1 is interpreted as the limit for V → ∞ of a well with finitely high
barriers. We now turn our attention to the stationary wave functions. First
of all we rewrite eq. 4.50 taking into account the quantization condition eq.
4.51:
m 1
G(x, x, En ) = 2
[2 + 2 cos 2(kn x − arctan γ)] 2i(kL−2 arctan γ)
=
i~ kn 1−e
  
m L nπ 1
= 2 4 cos2 kn (x − ) + (4.53)
i~ kn 2 2 1 − e2i(kL−2 arctan γ)
Where we did not substitute k = kn in the term outside the square parentheses,
in wait of taking the limit Eq. 4.37:

2 cos2 kn (x − L2 ) + nπ


2 2 (k 2 − kn2 )
|ψn (x)| = lim =
ikn k→kn 1 − e2i(kL−2 arctan γ)
 
2κn 2 L nπ
= cos kn (x − ) + (4.54)
2 + Lκn 2 2
q
where in the last passage κn = 2mV ~2
− kn2 . From eq. 4.54 the usual even and
odd wave functions inside the well can be obtained with the correct normal-
ization constant:
 q
2κ L


ψneven (x) = cos kn x −
q 2+Lκ 2
(4.55)
2κ L


ψ odd (x) = sin kn x −
n 2+Lκ 2

We now compute the wave functions in the external part of the well. The
easiest way to perform the calculation is to compute the Green function rep-
resenting the amplitude associated to a transition from a point xi inside the

70
4.6. One dimensional bound systems in piecewise constant potentials

Figure 4.5: Principal paths for each family in the inside-outside transition in
the finite square well potential.

well at a vanishingly small distance from the border, to a point xf outside the
well, as represented in Fig. 4.5. Since the problem is completely symmetric on
the two sides of the well, we only compute the part of the eigenfunctions on
the left part of the well. The Green function has the form:
−κ(x−L) 2i(kL−arctan γ)


m te 1 + e
G(L− , x, E) = 2 (4.56)
i~ k 1 − e2i(kL−2 arctan γ)
with t = 1 + r. For k → kn , eq. 4.56 can be rewritten as
−κ (x−L) 4 cos2 kn L
m e n
(
−
2
i~2 kn 1−e2i(kn L−2 arctan γ)
for n even
G(L , x, En ) = −κ n (x−L) 2 kn L (4.57)
m e 4 sin 2
i~2 kn 1−e2i(kn L−2 arctan γ)
for n odd

Now we compute ψn (L− )ψn∗ (X), for example for n odd (the calculation is
identical for n even)

e−κn (x−L) 2 kn L (k 2 − kn2 )

ψnodd (L− )ψn∗ (x) = 4 sin lim =
ikn 2 k→kn 1 − e2i(kL−2 arctan γ)
2κn kn L
= e−κn (x−L) sin2 (4.58)
2 + Lκn 2
And dividing by ψn (L− ) from eqs. 4.55 we obtain for x > L
 q
2κ
cos kn L2 e−κn (x−L)


ψneven (x) =
2+Lκ
q
−κ (x−L) (4.59)
ψ odd (x) = 2κ L
n 2+Lκ
sin kn 2
e n

Notice that the wave functions obtained are continuous, differentiable, and
with the correct normalization factor since the transmission and reflection co-
efficients automatically take into account conservation of flux at interfaces.

71
 4. The sum over paths approach

4.6.3 Asymmetric infinite square well

In order to demonstrate the generality of the method in approaching piecewise
constant potentials, we derive the quantization condition for the asymmetric
infinite square well or “step in a box” potential [347, 348]:


 ∞ if x< 0

0 if 0<x<L
V (x) = (4.60)


 V if L≤x≤L+M
∞

if x> L+M

Notwithstanding the conceptual simplicity of the method, when the number

of sections at constant potential to be dealt with is increased, the number of
paths to consider grows exponentially. As a consequence, for complicated piece-
wise constant potentials calculations may become quite burdensome. However,
in this case all the paths can still be counted with relative ease. In Fig. 4.6
the basic paths joining two points xi and xf in the left side of the asymmetric
potential are shown, divided in two categories A and B.
It can be shown that, although the problem is asymmetric, the paths have

Figure 4.6: Principal paths for the asymmetric infinite square well potential.

the same mathematical expression regardless of whether xf is to the left or

the right of xi . As can be seen from Fig. 4.6 , the difference between the
categories of basic paths “A” and “B” is that the former only contains paths
that stay within the left part of the barrier. Thus for paths of the category

72
4.6. One dimensional bound systems in piecewise constant potentials

“A” at least one transmission to the right part of the barrier must take place
before periodic paths in the right side of the barrier can be added to the total
sum over paths. On the contrary, for paths of the category “B” all periodic
orbits, either of the full well or of the left or right side must be added right
from the start. By defining, with the same notations of Fig. 4.6,

GA = GA1 + GA2 + GA3 + GA4 GB = GB1 + GB2 (4.61)

we can write the total Green function for the left part of the well as

 !p+1 !p 
∞
X ∞
X ∞
X
G = GA r02p (t12 t21 )p e2ip(k1 L+k2 M ) (r0 r11 )n e2ik1 Ln (r0 r22 )q e2ik2 M q +
p=0 n=0 q=0
 !p+1 !p+1 
∞
X ∞
X ∞
X
+GB r02p (t12 t21 )p e2ip(k1 L+k2 M ) (r0 r11 )n e2ik1 Ln (r0 r22 )q e2ik2 M q 
p=0 n=0 q=0
(4.62)

By substituting the formulas for reflection and transmission coefficients

k1 − k2
r0 = e−iπ r11 = = −r22 t12 = 1 + r11 t21 = 1 − r11 (4.63)
k1 + k2

and evaluating G for xi = xf we finally find an expression for G(x, x, E) in the

left part of the well

2m sin(k1 x)(k1 sin k2 M cos k1 (L − x) + k2 cos k2 M sin k1 (L − x))

G(x, x, E) = −
~2 k1 k2 sin k1 L cos k2 M + k1 cos k1 L sin k2 M
(4.64)

Leaving aside the computation of eigenfunctions, we immediately see from Eq.

4.64 that the poles of G(x, x, E) correspond to

k1 tan k2 M = −k2 tan k1 L (4.65)

which is the correct quantization condition for the “step in a box” potential for
both real and imaginary values of k2 [347, 348].
The condition in Eq. 4.65 cannot be expressed in terms of the generalized
WKB formula Eq. 4.39, and cannot be obtained using semiclassical schemes.
Thus, we believe that this result effectively demonstrates that the method
we use for piecewise constant potentials is not approximate or semiclassical,
provided all paths from xi to xf are taken into account.

73
 4. The sum over paths approach

4.7 One dimensional open system: square bar-

rier tunneling
We consider the problem of the tunneling of a quantum particle through a
square potential barrier, precisely a potential of the form:
(
0 if x < 0 ∨ x > a
V (x) = (4.66)
V0 if 0 ≤ x ≤ a

Our goal is to find the transmission probability as a function of energy T (E)

for an incoming particle from the left side of the barrier, which is the square
modulus of the transition amplitude t(E). In our approach T (E) can be defined
as
1
T (E) = |t(E)|2 = |G(0− , a+, E)|2 (4.67)
|G0 |2
Where G0 is the G(0− , 0− , E) Green function for an unreflected particle, i.e.
G0 = i~m2 k so that, infact, T (E) is simply |G(0− , a+, E)|2 without any pre-
factor.
The relevant coefficients for this problem are

k2 − k1 2k1 2k2
r21 = t12 = t21 = (4.68)
k1 + k2 k1 + k2 k1 + k2
√ p
with k1 = ~1 2mE and k2 = ~1 2m(E − V ). Considering all the infinitely
many non classical paths performing an arbitrary number of internal reflections
in the barrier, the desired transmission amplitude can be written as
∞
X 4k1 k2
t(E) = t12 t21 (r22 )2n ei(2n+1)k2 a = (4.69)
n=0
(k1 + k2 )2 eiak2− (k1 − k2 )2 e−iak2

For E ≥ V0 the wavenumber k2 is real; in this case we obtain the transmission

coefficient
1 1
T (E) = |t(E)|2 = 1 k1 k2 2 2
= V02
(4.70)
1+ (
4 k2
+ k1
) sin (k2 a) 1+ sin2 (k2 a)
4E(E−V0 )

On the contrary, √
for E < V0 , k2 is imaginary. In this case, using the substitu-
tion iκ2 = k2 = i V0 − E we have

1
T (E) = V02
(4.71)
1+ 4E(V0 −E)
sinh2 (κ2 a)

Thus we obtain the known results for the transmission (and reflection) coeffi-
cients for a rectangular potential barrier.

74
4.8. One dimensional smooth potential: the harmonic oscillator

4.8 One dimensional smooth potential: the har-

monic oscillator
Up to here we have only analyzed piecewise constant potentials, for which
the approach produces exact results. Since all continuous potentials can be
approximated by piecewise constant ones, it is tempting to try extending the
procedure to the general case. More explicitly, this means using the elementary
propagator Eq. 4.41 and evaluating transmission and reflection coefficients for
the continuous potential.
It is to be expected that the result will be approximate: indeed, we always
have to keep in mind the analogy with optics, where the problem of finding
transmission and reflection coefficients from a film with continuously varying
refraction index has analytical solutions in very few cases [349, 350]. It is
perhaps less evident that the result, using the simplest possible discretization
scheme for evaluating the coefficients, will reproduce the results of the Wenzel-
Kramers-Brillouin approximation (WKB).
As an example we now consider the case of a quantum object confined in a
harmonic potential:
1
V (x) = mω 2 x2 (4.72)
2
First of all we evaluate the reflection and transmission coefficients at the classi-
cal turning point, by computing the limit of r and t for a continuous potential
across the point xCT for which V (xCT ) = E:
 − i π 2 2 √ π
rCT = lim = e−i 2 tCT = lim = = 2e−i 4 (4.73)
→0  + i →0  + i 1+i
Since V 0 (x) is always different from zero at the turning point. The first of the
two limits in Eq. 4.73 corresponds to the − π2 phase shift for the reflection for
a soft (continuous) potential which is assumed in the Van Vleck-Gutzwiller
propagator Eq. 4.38 and in the WKB approximation [313]. These coefficients
are sufficient for computing the spectrum of energy levels: the action S0 (E)
for a single path running a full periodic orbit in the harmonic oscillator is the
well known result
Z q 2E2 s  
mω 1 2 2
2πE
S0 (E) = 2 q 2m E − mω x = (4.74)
− 2E
2
2 ω
mω

Thus
P∞ all2pGreen functions will contain a factor consisting of the geometric series
i 1
S (E)
p=0 r e ~ 0 = i( 2πE −π )
. The poles of a generic Green function will
1−e ~ω
correspond to
 
2πE 1
− π = 2nπ → En = n + ~ω n = 0, 1, 2, ... (4.75)
~ω 2
To proceed further, we need to compute a cumulative transmission coeffi-
cient tcum (x) for paths going from point xi = 0 to xf = x, and we approximate

75
 4. The sum over paths approach

a smooth potential by a piecewise constant (staircase) one. We discretize the x

coordinate in points x1 , x2 , x3 , ... with xi+1 − xi =  and we define the potential
(i) (i)
at point xi+1 as V (i+1) = V (i) + Vx , where Vx is the discrete derivative

V (i+1) − V (i)
Vx(i) = (4.76)

By doing so, all the transmission coefficients from the staircase steps have a
similar form:
2ki 2 2
ti = = q ' (i)
(4.77)
ki + ki+1 (i)
Vx 2− Vx
1+ 1− E−V (i) 2(E−V (i) )

where in the approximate equality we used first order Taylor expansion around
 = 0. Note that the expansion in Eq. 4.77 is valid for all discretized inter-
vals except the one containing the classical turning point, for which necessarily
(i)
Vx > E − V i and the first order approximation is not meaningful. This jus-
tifies a separate treatment of tCT as in Eq. 4.73. The cumulative transmission
factor tcum (xn ) up to the point x = xn can thus be evaluated by multiplying
all the ti , i.e.
n
Y 2n
tcum (xn ) = ti = (i) (i) (j)
Vx Vx Vx
P P
i=1 2n − 2n−2  i E−V (i) + 2n−4 2 i6=j (E−V (i) )(E−V (j) ) + ...
(4.78)
By returning to the continuous case, Eq. 4.78 becomes
1
tcum (x) = Rx V 0 (x0 ) Rx V 0 (x0 ) R x0 V 0 (x00 )
(4.79)
1 1
1− 4 0 E−V (x0 )
dx0 + 16 0 E−V (x0 ) 0 E−V (x00 )
dx00 dx0 + ...

Computing the above expression explicitly gives

1
tcum (x) = =
1 + 14 log | V E(x) − 1| + 1
16
log2 | V E(x) − 1| + 1
48
log3 | V E(x) − 1| + ...
1
=P h in (4.80)
∞ 1 V (x)
n=0 22n n!
log | E
− 1|

The above sum can be performed exactly, producing the result

 − 1 s  
 V (x)  4  ki 
tcum (x) =  − 1 =   (4.81)
E kf

In a similar way, it is possible to show that tcum (x) needs to be taken into
account only for the non-periodic part of the paths since, for paths going back
and forth and returning to the initial point, the total transmission coefficient
along the path vanishes in the limit  → 0. Finally, reflection coefficients in

76
4.8. One dimensional smooth potential: the harmonic oscillator

the same limit vanish everywhere, except at the classical turning point. These
results have also been confirmed through numerical computations.
π
= e−i 2 , and Eq. 4.81
The results in Eq. 4.73 for the reflection coefficient rCT r
 
for the cumulative transmission coefficient tcum (x) =  kkfi  do not depend
 

on the particular analytic form of the confining potential, as long as it is

continuous, differentiable, and U-shaped. If these factors are incorporated into
our elementary propagator Eq. 4.41, for all xi and xf in the classically allowed
region it can be written as
m i( ~1 S(xi ,xf ,E)−µ π2 )
G(xi , xf , E) = p e (4.82)
i~2 ki kf

i.e. identical to the Van Vleck-Gutzwiller propagator. It is now no surprise

(as it is proven in Schulman’s book [313]) that our results will in this case
reproduce those of the WKB approximation.
To find the bound state wave functions, the general procedure is to first find
their value at x = 0, then in the rest of classically allowed region using the same
paths as in Fig. 4.4, and finally in the classically forbidden region using the
paths in Fig. 4.5. We omit the details of the calculations, which are reported
in [340]. In the classically allowed region we find
s
2mω 2
|ψn (x)|2 = |tcum (x)|2 ·
En π 2
  r 
2 nπ xp 1 −1 mω
· cos + 2m (En − V (x)) + (n + ) sin x (4.83)
2 2~ 2 (2n + 1)~

In the classically forbidden region we have

s
mω 2
|ψn (x)|2 = |tcum (x)|2 ·
2En π 2
s s !2n+1
mω 2 mω 2 x2 √
− x~ 2m(V (x)−En )
· |x| + −1 + 2e (4.84)
2En 2En

We plot the probability density for the n = 3 eigenfunction for tcum (x) = 1
 − 1
 V (x)  4
and tcum (x) =  E − 1 . The wave functions obtained for tcum (x) = 1 are
continuous and differentiable (the left and right limits for the first derivative
are both zero) at the classical turning point; however, they do not reproduce
well the behaviour of the exact eigenfunctions away from the center of the
well, because all their maxima have the same height. The wave functions for
 − 1
 4
tcum (x) =  V E(x) − 1 agree with the WKB ones before the usually employed

correction through Airy functions.

77
 4. The sum over paths approach

Figure 4.7: Computed densities for n = 3. |ψ3 (x)|2 is exact, |ψ̃3 (x)|2 corre-
sponds to tcum (x) = 1 in Eqs. 4.83 and 4.84 while |ψ3W KB (x)|2 corresponds to
 − 1
 V (x)  4
tcum (x) =  E − 1 . The approximate densities are plotted as computed
without having been normalized

4.8.1 Comments on the results and their educational

translation
We may say that the method developed in this section extends the general
approach used by Feynman in QED for light (dealing with time independent
problems first, and introducing time in the description later) to the treatment
of confined systems for massive quantum objects. Indeed, due to the exact
correspondence between the Helmholtz equation and the time independent
Schrödinger equation, the method developed here is precisely the counterpart
of the stationary sum over path method for the photon discussed in Section
4.2.1. For piecewise constant potentials, the method gives exact results; while
for smooth potentials, using a discretization scheme to compute approximate
reflection and transmission coefficients, the Van Vleck-Gutzwiller energy de-
pendent propagator is recovered, and results agree with those of WKB. Al-
though not shown in this paper, the method is also extremely useful for the
educational treatment of partially confining systems (See Sec. 5.4.8), for which
it produces exact results as long as the potentials considered are piecewise con-
stant.
The approach uses a sum over paths perspective, starting from an elementary
propagator which takes into account transmission and reflection coefficients
from discontinuities in the potential encountered along the way, and summing
its contributions over all possible paths at fixed energy from xi to xf . The
paths considered pass through both classically allowed and forbidden regions;
in this sense, the approach is not “semiclassical”, as it takes into account the
property of the quantum object to overcome the limits imposed by classical
physics.
In the educational translation of this approach, the single path propagator
G(xi , xf , E) is explained as the basic “rotating arrow” [5] (phasor) associated

78
4.9. Sum over paths as an interpretative language for modern experiments

to each path, while the Green function G(xi , xf , E) given from the sum of
the G(xi , xf , E) phasors from all the possible paths, represents the amplitude
associated with a transition of the quantum object from xi to xf at fixed en-
ergy E. The probability of detecting the particle at xf is then proportional to
|G(xi , xf , E)|2 . The eigenvalues can be understood as those values of energy
which bring all the arrows in phase; and for all other values of energy the prob-
ability of detecting the quantum object is zero in the limit that the number of
paths considered goes to infinity [19].
For some of the cases presented, such as the infinite square potential well, cal-
culations can be carried out explicitly even at an elementary level; for others,
interactive simulations can be used to help students grasp the main conceptual
aspects without resorting to advanced mathematical techniques.

4.9 Sum over paths as an interpretative lan-

guage for modern experiments
From the 1980s on, a majority of experiments exploring and testing concep-
tual and foundational issues in quantum physics, such as wave-particle duality
[351, 352], non destructive measurements [353], indistinguishability [296, 354],
non locality and entanglement [355, 356] and so on, have come from the field of
quantum optics. In the discussion of these experiments, the language of Feyn-
man paths or processes [18] plays a fundamental role in guiding scientists in
the definition of the conceptual structure of the experiment before performing
precise calculations. This is understandable because quantum optics experi-
ments are, by large, experiments of interferometry; as such, the principal paths
for each photon in the interferometer arms are well defined, and correspond to
classical rays; while bifurcation of paths typically happens at beam splitters.
Furthermore, such experiments are almost always strictly based the source-
to-detector philosophy, in the sense that the photon source(s) and detector(s)
involved in the setup are clearly defined. So, in order to understand the struc-
ture of the experiment, it is sufficient to enumerate all the possible processes
that could lead from initial state A to final state B, limiting the analysis to the
possible classical rays including bifurcations at beam splitters. Transferring to
the high school context, at least in part, the way Feynman’s approach is used
in modern day quantum optics has been one of the primary goals of our work
(research question Q1).
It may be useful to restate the Feynman general rules for amplitudes and prob-
abilities (Section 4.1.1) in the form as they are formulated in one of the most
widely used textbooks in quantum optics [18]:

1. The probability of an event in an ideal experiment is given by the square of

the absolute value of a complex number A which is called the probability
amplitude.
P = |A|2 (4.85)

79
 4. The sum over paths approach

2. When an event can occur in several alternative ways, the probability am-
plitude for the event is the sum of the probability amplitudes for each way
considered separately; i.e. there is interference between the alternatives:

A = A1 + A2 P = |A1 + A2 |2 (4.86)

3. If an experiment is performed which is capable of determining whether one

or another alternative is actually taken, the probability of the event is the
sum of the probabilities for each alternative. In this case,

P = P1 + P 2 (4.87)

and there is no interference between alternatives.

In applying rule (2) it is essential to be sure that the situation described in rule
(3) is excluded. This means that the experimental arrangement must be such
that it is impossible, even in principle, to determine which of the alternatives
actually occurs.
Rules (2) and (3) constitute a particularly sharp formulation of the idea of
“wave particle duality”. This has suggested to us a possible way for over-
coming the well known conceptual difficulties originating from such concept
[77, 124, 142], and making it understandable and, more than all, verbalizable
for students. What we do is lead students to operate a semantic shift for the
expression “possible paths”. Initially, the possible paths are simply defined
as those compatible with the constraints imposed on the system; but starting
from the discussion of interaction free measurements [18] (see Section 5.2.5) we
redefine them as all the paths compatible with physical constraint and with
information that is acquired or otherwise available about the system in the
given experiment. As it will be seen clearly from our results in Chapter 6, this
strategy allows students to build an internal model of the otherwise vague con-
cept of “wave particle duality”, and to tell a story about it: acquiring “which
way” information about a system reduces the possible paths to only those com-
patible with it, destroying interference. If necessary, the discussion of the case
of imperfect detectors (treated by Feynman on p.82 of QED [4]) can then be
added. Although the sum over paths approach is not specifically tied to any
interpretation of quantum mechanics, this way of seeing things is intimately
connected to modern informational based foundational views [357, 358].
The generalization of the sum over paths to the sum of all histories, from an
initial to a final state, also allows to consider setups in which processes appear
that only differ by the exchange of two identical particles. From an experimen-
tal point of view, the simplest realization of such idea is the Hong-Ou-Mandel
effect, which has been experimentally verified for both photons [296] and elec-
trons [359]. In our educational path we will fully exploit this route to introduce
the difference between bosons and fermions, and the Pauli exclusion principle.

80
Chapter 5
The design of our proposal

The structure of the present chapter is the most complex among all the ones in
this dissertation, so it is worthwhile to start with a foreword about its structure
and rationale. Its overall objective is to provide an overview of the chronolog-
ical development of our learning path, which we feel is essential for allowing
the reader to understand the results obtained in the different experimentations
reported in the following Chapter 6; however, we also wish to give the reader
an unitary view of the main directions followed in the development of the se-
quence. Thus, the chapter is organized as follows: in Section 5.1 we discuss the
most fundamental educational choices on which our design was based, report-
ing also, where pertinent, how such choices were adjusted while proceeding the
experimentations. This can be considered as an expanded view, presented in
a less schematic fashion, of most of the threads which were identified in the
Introduction (Section 1.2) in the context of defining our research questions.
In Sections 5.2 and 5.3 we provide detailed accounts of the structure of the
central core of the sequence, as it was used in the tests of 2014, and 2015 re-
spectively. In Section 5.4 we examine most of the GeoGebra simulations which
were used in the tests, grouping them by similar themes. Finally, in Section
5.5 we describe the process of constructing a wider learning environment for
our teaching-learning sequence in view of the test in high school. Such process
in general did not touch the main core of the sequence, which, apart from a
couple of more advanced subjects which were left out1 , remained as presented
in Section 5.3.

1
The few differences between the structure of the main core of the sequence for the two
2015 tests with teachers and high school students are also specified in Section 5.3.

81
 5. The design of our proposal

5.1 Main educational choices

5.1.1 Making the conceptual and epistemological struc-
ture of the theory transparent.
An element often highlighted in the literature is the difficulty for students to ac-
cept concepts and predictions of quantum physics which are very distant from
the classical paradigm [32]. This has been connected to the necessity of ex-
plicitly discussing with students epistemological and foundational perspectives,
with the non secondary goal of leading students to take a personal, individual
but socially constructed position on these subjects, favoring appropriation and
the incorporation of quantum physics into the student’s cultural identity [34].
It has also been shown that in courses where the instructor choses not to take
any particular position concerning foundational and interpretative aspects of
quantum theory, making of them a sort of hidden curriculum [360, 361] it
is more likely that students maintain conceptions shaped by classical realism
[125].
A distinguishing feature of Feynman’s formulation is its conceptual clarity, and
the possibility it offers to tell a story, albeit very different from the classical
one, about the evolution of quantum systems. Feynman’s approach allows stu-
dents to verbalize explicitly the content of problematic concepts of quantum
physics, such as wave particle duality, the impossibility of assigning a trajec-
tory to the quantum object, the role of measurement, and helps them build
consistent, and as such satisfactory, internal models.
We start considering the example of wave particle duality as it is probably the
point on which, based on our results, the effectiveness of our approach is more
clearly evident. In Ref. [134], the authors define the first of the two main
research questions of their work as:
”How do students of quantum mechanics conceptualize the mental models in-
volved in the wave-particle paradox?”
In the sum over paths approach, it is the model itself that tells teachers how
students should do that. We may say that the Feynman formulation provides
the ideal setting in which Lévy-Leblond’s suggestion not to talk of waves and
particles, but only of quantum objects having specific properties can be real-
ized. Using such approach, in fact, the primary rule for the quantum object
is “follow all possible paths”, where the term possible must be interpreted as
“compatible with available information about the system”. Wave particle du-
ality can then be expressed verbally as the reduction of possible paths due to
information acquired through a measurement, or by any other means avail-
able. It is such reduction which, by negating interference between alternative
possibilities, produces a particle-like behaviour of the quantum object. We
hypothesized that such description would have been more easily acceptable for
students; and would have led them to build stabler, and more explanatory,
mental models than those based on the concept of wave function collapse. As
we will see in Chapter 6, the results of our experimentation have convincingly

82
5.1. Main educational choices

confirmed such conjecture.

In our educational path, we gradually, passing through an analysis of modern
experiments, lead students from such description of wave-particle duality for
problems involving only a single quantum object, to construct a representation
of experimental setups in which all alternative, indistinguishable histories lead-
ing from the same initial state, to an identical final one, can interfere, while for
distinguishable processes the classical probability rules are used [18]. This, we
believe, provides students with a transparent view of the deep conceptual and
epistemological meaning of quantum physics: all the alternative ways in which
the same thing can happen, which we cannot distinguish even in principle, do
not realize alternatively with some probability, but are lumped together, in an
inextricable whole, producing interference.
The entire process of gradually leading students to build such model may be
seen as an example of the importance of providing explicit assistance to stu-
dents in the formation of new ontological categories [83, 370]. It also should
be apparent, from the above formulations, that we take a definite perspective
on the foundational issues of quantum physics, precisely the one assigning a
central role to the concept of information [357, 358, 362, 363].
Another feature of the Feynman formulation which is worth considering is
that it makes the classical limit completely transparent, allowing an easy and
convincing derivation of the classical laws from the rules valid for quantum
objects. This facilitates the task of explaining to students the emergence of
the classical models in certain conditions, rather than only trying to replace
the classical framework with the quantum one [83].

5.1.2 Discussing modern experiments

A central theme which we have highlighted in Sec.1.2 concerns the tension
between the necessity, which educators often perceive, to explain the historical
crisis which led to the emergence of the quantum paradigm, and the goal of
leading students to build entirely new conceptual schemes to interpret quan-
tum phenomena [76, 167]. In fact, research has shown that the semiclassical
models of old quantum theory, which are often introduced first in a histori-
cal approach, tend to settle in the minds of students, leading them mix the
conceptual frameworks of classical and quantum physics, developing a hybrid
worldview that contains elements of both [23, 24, 25, 118]. In other terms,
rather than constructing a genuine quantum worldview, students tend to a
progressive incorporation of new quantum ideas in the old classical framework.
Concerning the perceived necessity of back tracing the roots of the rupture
of the classical paradigm, it is also useful to keep in mind the criticism by
Nersessian [364] to the general applicability of the Kuhnian concept of scien-
tific revolution [79]. According to Nersessian, a change of scientific paradigm
may revolutionise, in a short time span, the objects and instruments which
scientists use to analyze reality; however it is not necessarily the case that the
meaning which scientists attach to such entities appears in its definitive form

83
 5. The design of our proposal

immediately after the perceived paradigm change. Instead, evolving meaning

schemas may be produced, in which the interpretation of objects and structures
of the new theory is initially mediated by concepts proper of the old paradigm,
and only gradually, with time, their meaning settles to a definitive form. The
interpretation of the uncertainty principle, and the reasons behind the special
status of measurement in the theory, can be seen as a exemplary case in quan-
tum physics. As shown in the famous debate between Bohr and Heisenberg,
in the first years following the quantum revolution a widespread conception
was that the uncertainty principle was due to a physical perturbation of the
observed system due to the interaction with the measurement apparatus. To-
day, due also to modern experimental evidence which is discussed within our
proposed educational path, the prevailing idea in the scientific community is
that the uncertainty principle expresses an intrinsic limitation on information
available on the system, and as a consequence on its state2 [365, 366]. Since
the interpretation of concepts, in the community of physicists, often evolves as
a response to novel experimental evidence, exploring known principles under
new perspectives, this may be a first argument in favour of presenting modern
experiments early in an introductory course.
On a different, but related level, taking into account Vosniadou’s suggestions
for favoring a process of conceptual change [26, 28, 83], we argue that modern
experiments are often more effective in producing cognitive conflict on a local
scale, with the aim of preventing the formation of synthetic conceptions, at
specific steps in the learning process. Indeed, trying to achieve an interplay
between the introduction of theoretical concepts and the discussion of experi-
mental results which was effective in preventing students from forming hybrid
conceptions has been a major theme in our work (research question Q5). We
take as an example the point of our sequence in which students observe the
formation of the interference pattern, in a Young setup with one photon at a
time, through the gradual accumulation of spots resulting from the detection
of individual photon at the screen. The early introduction of this experiment
and the video of its realization are meant to counter a very popular synthetic
conception which students form after learning about the existence of photons:
that interference is due to photons interfering one with another. However, af-
ter observing the video students may form yet another hybrid conception to
reconcile what they observe with the classical paradigm: that the photon is a
somewhat extended object, and that interference happens because the photon
2
Actually, in the recent literature, the original “Heisenberg principle” is increasingly being
separated into two very different concepts [366, 367]. The name “uncertainty principle” is
being reserved for the relations between non commuting observables derived from Robert-
son’s theorem [368], which express a limitation on the preparation of a system, and as such
have nothing at all to do with measurement. The name “Heisenberg principle”, instead,
is nowadays often used to refer specifically to information-disturbance trade-offs, which do
appear in quantum measurements, at least in certain conditions, and whose origin is similar
in spirit to the original example of Heisenberg microscope. Although we did not adopt this
approach or nomenclature in our work, this is certainly a tendency worth considering in the
future.

84
5.1. Main educational choices

physically divides in two parts at the slits. In order to counter the formation
of this specific conception, we introduce at this point the Grangier [295] ex-
periment of 1986 demonstrating photon indivisibility. Note that, previously,
only an experiment by Clauser in 1974 [369] had given direct evidence of the
photon being an indivisible entity, and Grangier’s one is by far the most direct
and conclusive [18].
As a final remark, we remind Chinn and Brewer’s [370] general suggestions for
presenting, in science instruction, experimental evidence as a cognitive conflict
strategy. According to the authors, data in conflict with the old theory pre-
sented to students should be as unambiguous and credible as possible; should
come from multiple experimental situations, and should try to address as many
sources of ambiguity as possible. From this point of view, it appears evident
that in the context of quantum theory the most unambiguous and credible ex-
perimental data on some particularly controversial aspects comes from modern
experiments.
With these considerations in mind, we chose not to adopt an historical ap-
proach, but to start by introducing, alongside with some irrenuciable historical
experiments such as the photoelectric and Compton effects, relatively recent
historical results, sharply highlighting the non classical features of quantum
physics, and the meaning attached to concepts in the current understanding
of the theory by the scientific community. Among these, along with the one
by Grangier discussed above, the most important one for our educational path
were the Mach-Zehnder experiment with individual photons, the Zhou-Wang-
Mandel [354] test on measurement without interaction, the Hong-Ou-Mandel
[296] setup exploring the consequences of photon indistinguishability.

5.1.3 Addressing mathematical difficulties of students

The mathematical difficulties associated with the teaching of quantum physics,
which we reviewed in Sec. 3.1.2, are well known. Right from its origins, [4, 5]
one of the main motivations for the use of Feynman’s approach in physics
education were related to the possibility of using simpler mathematical tools.
From a formal point of view, such possibility is due to the fact that, rather
than finding solutions to the Schrödinger equation, Feynman’s method con-
structs the Green function for the same equation, representing it as a sum of
complex amplitudes computed over all possible paths. Since the Green func-
tion has its own physical meaning, which allows to interpret phenomena and to
independently define the concept of a “wave function”, the Schrödinger equa-
tion can be omitted, and with it all the difficulties associated with differential
equations, and with students’ understanding of waves. However, certainly not
all the mathematical difficulties associated with quantum physics can easily
be bypassed; in particular, difficulties related to students’ conceptions of the
classical and quantum probability rules, discrete and continuous distributions,
normalization, and uncertainties [187] certainly remain, and must be carefully
approached, especially by formulating the relevant concepts in a clear and un-

85
 5. The design of our proposal

ambiguous way, and by designing and discussing with students exercises and
problems specifically related to the probability aspects enumerated above 3 .
Indeed, although the importance of leading students to acquire a primarily
conceptual understanding of quantum theory cannot be underestimated, in de-
signing our proposal we took the clear stance that students, at the end of our
course, should have been able to solve exercises and problems at least as well as
students of other introductory courses in quantum physics. To reach this ob-
jective, we designed several introductory exercises specifically concerning the
sum over paths formulation; we translated the solution of other, traditional
problems to the language of Feynman paths; and we made sure to complete
all the steps to introduce those universal concepts of quantum theory that are
independent of the choice of a particular formulation; for example, once the
concepts of “wave function”, “stationary states” or “allowed energy levels” are
defined starting from the point of view of Feynman paths, they can then be
used independently to approach traditional exercises. In the version of the
sequence which was tested in high school, for example, students were required
to solve as home assignment, in addition to the specific problems accompany-
ing the sequence (Appendix B) all the traditional exercises contained in the
quantum physics chapters of their textbook (the Italian version of the widely
used book by Cutnell and Johnson [193, 194]).

5.1.4 Visualization and the use of simulations

Several researchers have pointed out the importance of visualization in learn-
ing and teaching and developing scientific understanding [181, 187, 302]. Feyn-
man’s formulation permits to visualize quantum processes, or more precisely
provides a pictorial representation of the underlying mathematical model. This
aspect is one of the main reasons underlying the scientific fortune of Feynman
diagrams, which are a graphical representation of the extension of the path in-
tegral algorithm to processes involving multiple interacting particles, exchanges
between identical quantum objects, creation and destruction processes.
As it was discussed in Section 4.3 interactive simulations play a central role
in all courses based on the sum over paths approach. In our case, simulations
were designed using the open source software GeoGebra [339], and used during
explanations, and in exploration and inquiry activities; they allow students to
familiarize with Feynman’s model both in simple situations and in more com-
plex contexts such as interferometers, bound systems, tunneling from barriers.
It has to be remarked again, however, that the issue of visualization in the
teaching of quantum physics is debated. There are two different objections,
which need to be considered separately. The first one concerns, in general, the
possibility of providing any form of pictorial representation of the quantum
object [32, 33], since any kind of visualization would be misleading. To use the
3
In the experimentation with high school students, the recent curriculum reform which
significantly expanded the time devoted to probability and statistics has also been helpful.

86
5.1. Main educational choices

words of physicist Paul Davies, “it is impossible to visualize a wave-particle,

so don’t try.” [371]. Although there has been research on devising the best
ways of producing simulations in which a quantum object is visualized in a non
misleading way [302, 303], we mostly agree with the content of the criticism.
In designing our simulations we tried as much as possible to avoid any inter-
ference with the classical paradigm (research question Q6), and in particular
we never provided any representation of the quantum object itself; instead, the
graphical interface shows the geometrical structure of the problem considered,
the source and detector(s), the possible paths, and the sum of vector ampli-
tudes producing detection probabilities. The second criticism is more subtle,
and is related specifically to the kind of visualization which is provided by the
sum over paths formalism [15]. It is argued, in essence, that using an approach
that is founded on the analysis of possible paths of a pointlike entity, students
could remain attached to the classical concept of trajectory, and believe that,
at the end of the story, the quantum object follows only one of the possible
paths. From our perspective, this criticism can be interpreted, again, as the
possible formation of an inconsistent, hybrid quantum-classical synthetic con-
ception; and as such, we have tried to counter it, right from the initial design of
our sequence, presenting convincing experimental evidence at the point in the
sequence at which such conception could be formed. Specifically, our approach
concerning this point is based on a detailed analysis, made also with the help
of a computer simulation, of the results of a Mach-Zehnder experiment with
one photon at a time, and the comparison of the full setup with the one in
which either one of the two arms of the apparatus is removed or blocked (see
Section 5.2.5). The resulting detection probabilities in the different cases, very
sharply show that the hypothesis that, in the full setup, the photon has taken
only one or the other of the two possible paths is untenable [137, 293]. The
result is then extended, beyond the particular setup, by considering the analo-
gous case of the two slits experiment, and comparing the resulting probability
distribution with the one corresponding to the cases that either one of the slits
is blocked. According to our results, this approach is efficient in countering
the appearance of the conception hypothesized in Ref. [15].
Research has pointed out that interactivity and the possibility for students of
exploring the simulated phenomenon by varying parameters is a necessary con-
dition for simulations to be productive in educational use [298, 299, 300]. Ge-
oGebra offers the possibility of designing highly interactive simulations through
the use of sliding bars ad checkboxes; allows to connect different representa-
tion in adjacent windows of the workspace, and is a very widely known and
used software in the community of mathematics and physics teachers. This,
in perspective, allows teachers to work autonomously on the simulations we
produced, modifying and adapting them to their educational needs.

87
 5. The design of our proposal

5.1.5 Pursuing appropriation

In the initial design of our sequence we thought of appropriation mainly as a
process in which the disciplinary content is internalized by the student in such
a way that it can be used, in a personal way, for interaction with others, with-
out borrowing the language, or concepts, from external sources. In this sense,
the main strategy we used for promoting appropriation (research question Q8)
was to make the conceptual and epistemological structure of the theory com-
prehensible and verbalizable by students, in order to help them build consistent
and satisfying mental models, which they could use as a source for their dis-
course on quantum physics, in place of external authorities such as the teacher,
or the textbook. And in this sense, the main indications for appropriation we
were looking for in our data were of the linguistic kind. Analyzing the re-
sults our first test in 2014, we found clear signs of such process happening:
in their answers to open questions of the final test, students frequently wrote
in first person, identifying themselves with the experimenter, or with the sys-
tem itself. They often used a colloquial language, producing personal terms to
express concepts, different from those which were proposed to them, but still
pertinent; and in general they expressed satisfaction with the internal models
they had built.
Based on these observations, when preparing for the 2015 test of our sequence
with high school students, we decided to take the route of designing a properly
complex learning environment to encourage student’s appropriation (research
question Q9), in the wider, and more precisely defined meaning, introduced in
Ref. [34]. From an educational and pedagogical point of view, the significance
of such choice is self-evident: the objective of supporting high school students
in the construction of their personal and cultural identities [75], and leading
them to appropriate of quantum physics as an element of their own episte-
mological discourse is certainly worthwhile to pursue. From the point of view
of research alone, furthermore, we thought it would have been interesting to
compare our results with students to those of the group of Bologna, who was
starting a similar, if larger in scale, work in high school on quantum physics
using a totally different educational reconstruction of the content. The steps
we took to integrate our educational path into a learning environment suitable
for this objective are described in Section 5.5.

5.1.6 Including a practical experimental dimension

The idea of introducing in the sequence hands-on experiments, to conduct
activities which were both engaging for students and meaningful from the
disciplinary point of view has always been in our mind (research question Q7).
However, in the first version, due in part to time constraints, we were able to
include only a very limited number of experimental activities, which mostly
concerned light interference and diffraction. For the revision of the sequence
which led to the 2015 tests, we made a specific work in this direction, and

88
5.2. The 2014 version

guided by the existing literature we introduced simple experiments meant to

observe and measure relevant quantities such as the Planck [372, 373], and
Rydberg constants [374], sometimes improving these activities in the sense of
being able to obtain more accurate results with very low-cost materials [59].

5.2 The 2014 version

5.2.1 Structure
The first version of our teaching sequence is schematically represented in Fig.
5.1. We provide here a brief summary of the main steps of the sequence (the
“spinal cord” of the graph in Fig. 5.1), and in the next subsections we provide
a more detailed description of the content and educational strategies for each
step.
The sequence starts from the phenomenology of wave optics, as observed in
the classical experiments of interference and diffraction. The sum over paths
method, which in this context may be more properly called the method of
paths and “phasors” [195] is first introduced as a convenient way of represent-
ing wave phenomena.
The next step is the discussion of experiments leading to the necessity of a
photon model: photoelectric effect, two slits interference with individual pho-
tons, and the Grangier experiment on photon indivisibility. Each experiment
is discussed with students in its own significance, carefully monitoring the de-
velopment of their ideas and conceptions, and finally the model of the photon
following all possible paths from source to detector is introduced as a consis-
tent solution to apparent paradoxes.
A central part of the sequence is devoted to discussing conceptual and episte-
mological themes through the analysis of modern quantum optics experiments
in the language of Feynman paths. Evidence discussed include the single pho-
ton Mach-Zehnder and Zhou-Wang-Mandel experiments, and the main threads
in this phase are a critical re-evaluation and refutation of the classical concept
of trajectory, a deeper inspection of the classical and quantum probability
rules, and the interpretation of wave particle duality from the sum over paths
perspective.
We then review some diffraction and interference experiments with massive
particles, an proceed to extend, by means of analogy, the formalism used for
photons to massive quantum objects.
In the next step we deal with the problem of the limits of classical mechanics
and geometrical optics, i.e. the correspondence principle, whose treatment is
particularly natural from the Feynman perspective [7, 336].
Finally, using the approach described in Section 4.6 we treat simple bound sys-
tems, such as the particle confined to a ring, and the infinite square well, and
demonstrate the existence of discrete energy levels in confined systems. This
leads us naturally to the discussion of the spectrum of the hydrogen atom, that

89
 5. The design of our proposal

concludes our sequence.

GeoGebra simulations (Sec. 5.4), which are used to sustain students’ com-
prehension both during lessons and in home activities, concern both simple
problems of interference, transmission, reflection and refraction of light seen
from a photon point of view, and modern interferometry experiments. Home
and classroom exercises (Appendix B) on the sum over paths formulation were
also proposed to students.

Figure 5.1: Schematic representation of the teaching sequence. In the cen-

tral pathway, the content presented is (to some extent artificially) divided in
steps. In blue squares, the experiments discussed at each step. In orange, the
interconnected conceptual and foundational themes discussed.

5.2.2 The sum over paths method in wave optics

The sum over paths method can initially be seen as a convenient way for de-
scribing interference phenomena in a classical wave perspective [11]. We discuss
the Huygens’ principle, whose idea of a wave producing new wave sources at
all points in space, has well known analogies with the path integral formula-
tion [324, 375] (Fig. 5.2 (a)). Through simple exercises students are led to
compute the value of the amplitude of a monochromatic wave at a given point

90
5.2. The 2014 version

P in space by summing the amplitude vectors resulting from all possible op-
tical paths that the disturbance in the wave could have followed to reach P
(provided it remained coherent) from the original source S. The importance
of the source-to-detector philosophy [6], which is often implicitly assumed in
the usual treatment of interference phenomena in wave optics, is strongly high-
lighted. In this phase, simple experiments are performed and discussed using
Tracker as a tool to produce a quantitative analysis of interference fringes (Fig.
5.2 (b))

(a) (b)

Figure 5.2: (a) Animation illustrating the connection between the Huygens
principle and the idea of using all possible optical paths from source to detector.
(b) Two slit interference pattern with superimposed Tracker analysis of the
light intensity.

5.2.3 The photon concept

The quantum discontinuity emerges in our approach from the photon idea.
As diffusely anticipated, we do not commit to a historical reconstruction, but
use experimental evidence gathered at different times, in the unifying perspec-
tive of providing a logically convincing, unambiguous construction of the idea
of photon and of its fundamental properties. In particular, we discuss the
following experiments:
ˆ The photoelectric effect, as basic evidence of granularity of light interacting
with matter, and introducing the Planck relation E = hν.

ˆ The Grangier et al. 1986 experiment [295] on photon indivisibility.

ˆ Two slit interference experiments with single photons [376] with a video
[377] of the accumulation of individual photons, introducing the probabilistic
interpretation.
Students have to be, first of all, given concrete evidence that the photon exists:
discussing traditional evidence (i.e. the photoelectric effect) may be efficient

91
 5. The design of our proposal

in this respect. The existence of the photon, however, implies that the over-
whelming evidence in favour of a wave model of light should be discussed in
terms of photons. For example, if the photon exists, then how is the interfer-
ence pattern in Young’s experiment formed? It is known in the literature that,
if the point is not addressed, students may form a hybrid conception of the
pattern appearing as the result of interactions, or interference, between one
photon and another. Thus we show students a video of the accumulation of
the two slit interference pattern using a very weak light source that emits one
photon at a time. The conclusion to which we would like to lead students is of
course that the photon interferes with itself; however, students may try to save
their classical world vision by hypothesizing that the photon actually splits in
two at the slits and then recombines. Thus, at this point, we discuss the Grang-
ier experiment on the indivisibility of the photon. Only after having presented
decisive evidence to counter the possible appearance of mixed classical quan-
tum conceptions, we introduce Feynman’s model of the photon following all
possible paths from the source to the detector, arguing that, notwithstanding
its prima facie absurdity, it can explain all the previously considered evidence
in a consistent way.

5.2.4 The Feynman photon model

A basic single-photon problem in quantum physics consists in having a source
placed at point S, emitting a photon which may or may not be collected at
a detector at point P . As Feynman did in QED (see Section 4.2.1) we use a
time-independent approach to treat this kind of physical settings, adding some
remarks which may be useful to better clarify the situation in an educational
setting.
In a time-dependent sum over paths perspective, it is necessary assign to the
photon the phase φ = kx − ωt, and sum the corresponding amplitudes carried
over all paths (and considering all possible energies) to the final space-time
detection point. However, meaningful interference effects ar only obtained if
a photon wave packet of of finite width (coherence length or time [293]) is
used. Coherence length for a single photon is defined, through the uncertainty
principle, as the uncertainty on the photon localization, and is proportional to
coherence time:
h hc
lc = ∆x = = = ctc (5.1)
∆p ∆E
Using wave packets of finite coherence length means in practice that only paths
whose length differs of less than lc can interfere. As the uncertainty ∆E in
the photon energy is reduced more and more, two things happen: 1) coherence
length and time become very large, so that, in essence, all paths from S to P
must be considered, independently of their length; and 2) all paths associated
with an energy significantly different from E all interfere away, and need not be
considered; and since for paths at the same energy the ωt term only produces

92
5.2. The 2014 version

an irrelevant overall phase, such term also can be discarded. In this way, when
∆E → 0 the prescriptions of the time-independent approach are recovered:
use all paths at fixed energy from S to P independently of travel time, and
use the phase corresponding to the classical abbreviated action φ = kx. Of
course, it can be checked directly that the time-independent approach using a
long coherence time, an the time dependent one give consistent results [325].
Seen from a different perspective, the time-independent approach consists in
using the single photon equivalent of a monochromatic wave, i.e. a photon with
infinite coherence time. The approximation is meaningful, since the coherence
time of a single, approximately monochromatic photon in realistic settings is
much greater than the length scales of the experiment. In terms of the un-
certainty principle, such a photon can be understood to have a large (ideally
infinite) uncertainty in its emission time. Thus, in the quantum perspective,
a monochromatic wave cane be seen as ensemble of identical photons, or even
a single photon, whose energy is precisely defined, and by consequence, which
has an infinite uncertainty in space and time localization.
Although these concepts cannot be made explicit to students at the beginning
of the sequence, because the uncertainty principle has not yet been introduced,
we return to this concept later, analyzing it in retrospective. We believe that
a full clarification of this issue with students to have a high educational value,
as it establishes the role of the uncertainty principle in connecting the time-
dependent and time-independent descriptions of quantum physics.
A key difference with the classical perspective in wave optics lies in the prob-
abilistic interpretation. The square modulus of the resulting arrow at point
P must now be interpreted as (proportional to) the probability of detecting
the photon at P . A clear and precise formulation of this point is very impor-
tant, as students can be tempted to assign probabilities to individual paths; or,
they may remain confused with the issue of normalization. Also, the difference
between the classical and quantum ways of computing probabilities must be
clarified from the start. This can be done by comparing the resulting probabil-
ity distribution for a Young experiment with slits of finite width, with the same
setup when either one of the two slits are blocked. This last example can be
compared to the classical expectation for the case of shooting marbles against
a wall with two slits. Simulations have here an essential role in clarifying these
points. Finally, the “wave function” of a quantum object can be defined (again,
paying attention to the normalization problem) as the set of complex values
corresponding to the resulting amplitudes at all points in space [5].

5.2.5 Using the Feynman approach for treating concep-

tual and foundational issues
The discussion of epistemological matters in terms of Feynman paths, which
represents the core of our proposal, comprises three main elements:

ˆ The single photon Mach-Zehnder (MZ) interferometer [295] demonstrating

93
 5. The design of our proposal

the impossibility to attribute a single path to a quantum particle [137].

ˆ Diffraction of a photon from a single slit with a variable width, introducing

a discussion of the uncertainty principle [196, 197].

ˆ The Zhou-Wang-Mandel (ZWM) 1991 [354] experiment used to highlightthe

role of indistinguishability and which way information.
The Mach-Zehnder interferometer has been used by several authors in educa-
tion [137, 286, 287]. In the case of our sequence, its primary function is to
provide a convincing proof of the untenability of the classical trajectory con-
cept. We do not here describe the experiment in detail, as it will be more
thoroughly explained in Section 5.4, where the related simulation is presented.
Comparing the outcomes corresponding to different setups of the apparatus
proves that the photon cannot be thought as taking either one or the other
of two possible ways, but must be imagined as taking both of them simul-
taneously. In other words, it is impossible to associate a single path to a
quantum object. Exercises on the Mach-Zehnder setup (see Appendix B) are
also very useful for introducing to students the concept of normalization of
wave functions, and for improving their understanding of the quantum rule for
probabilities.
The phenomenon of single slit diffraction with a variable slit width is discussed
as a simple but effective introduction to the uncertainty principle [196, 197].
The diffraction pattern can be easily related to the momentum probability
distribution of the diffracted photon, and as the slit width is varied in a sim-
ulation, an intuitive understanding of the meaning of the principle can be
provided. Simple calculations lead to the approximate uncertainty relation
∆x∆px & h2 .
We pay particular attention to making clear that the uncertainty principle
should be interpreted as an intrinsic limitation to the information about com-
plementary variables which quantum states can contain, and not as a perturba-
tion of the system due by measurement, as in the old “Heisenberg microscope”
view. Indeed, we here first introduce, although we do not insist on it, the idea
that the quantum concept of “state” can be defined as the information which
is available on the system [365].
The ZWM apparatus is a two way, single photon interference setup where
“which way” information is collected in a non-destructive manner through a
clever use of nonlinear crystals. Since this experiment plays a pivotal role in
our sequence, we describe it in detail, along with its educational significance
for our learning path, in the following Section 5.2.5.1.
Overall, this part of the teaching-learning sequence, devoted to the analysis
of modern experiments, opens up the way for a collective discussion on the
profound epistemological fracture between classical and quantum physics, and
on current foundational views [358] in which information plays a crucial role.
Advancing the foundational debate with students to the level of its present
state of the art is an important thread in our proposal.

94
5.2. The 2014 version

5.2.5.1 The Zhou-Wang-Mandel experiment

The ZWM apparatus, in the two different setups which are used in the exper-
iment, is reported in Fig. 5.3: light emitted from a low intensity laser source
passes through a beam splitter, where it is divided into two beams, each one
of which excites a non linear crystal, capable of emitting a couple of photons,
with different frequencies and in different directions, called the “signal” and
“idler” photons. Because of the low intensity of the source, and of the low
efficiency of lithium iodate crystals, statistically only one of the two at a time
emits a couple of photons.
In the configuration shown in Fig. 5.3 (a), the apparatus is aligned in such a
way that the idler photon is directed towards detector B, regardless of which
one of the two crystals has emitted the couple. In this way, the events “the
couple is emitted by crystal 1” and “the couple is emitted by crystal 2” are
indistinguishable, and consequently, interfere: if the optical path length of one
of the two possible paths for the signal photon is varied, an interference pattern
is observed at detector A (Fig. 5.4).
If, on the other hand, the apparatus is aligned in such a way that in only
one of the two possible cases (emission of the couple from crystal 1 or 2) the
idler photon will arrive to detector B (Fig. 5.3 (b)), then the two cases are
no longer indistinguishable; the state of the system contains information on
which one of the two crystals has emitted the couple (in an abstract sense,
“which way” information has been obtained) and varying the optical length of
one of the possible paths of the signal photon no longer has any effect: the
photon count at detector A remains constant (Fig. 5.4). The main result
of the ZWM experiment consists in proving that the modification in the final
outcome of an experiment due to an intermediate measurement, which is pe-
culiar to quantum physics, should not be thought in terms of a disturbance,
but of information acquired or recorded about the system. The language of
the Feynman approach is especially appropriate for summing up the lessons
that can be drawn from the experiment. In particular it is sufficient to rein-
terpret the expression “all possible path” to mean “all paths compatible with
the information about the system”, to reconstruct the idea of wave function
collapse caused by a measurement, even if of non-destructive nature.
Since, from an abstract point of view, the ZWM experiment s nothing else
than a two way interference setup, it is useful at this point to go back with
students to the two slit experiment and discuss the example usually presented
concerning wave particle duality, i.e. the fact that a which way measurement
destroys interference on the screen. The discussion on the ZWM experiment
will help students to understand that the origin of the disappearance of the
interference pattern does not lie in the fact that one of the two slits has been
“blocked”, or in a physical interaction between the apparatus and the quan-
tum object, but must be found in the reduction of possible paths, due to the
acquisition of additional information on the system.
The ZWM experiment also allows to introduce a generalization of the concept

95
 5. The design of our proposal

(a)

(b)

Figure 5.3: Schematic representation of the Zhou-Wang-Mandel experiment.

(a) Configuration in which the two events are indistinguishable, and an inter-
ference pattern is observed varying the optical path of one of the two arms of
the interferometer. (b) Configuration in which the events are distinguishable,
and the interference pattern disappears.

Figure 5.4: Experimental results for the ZWM experiment in the indistinguish-
able and distinguishable cases (reproduced from [354])

of path. In fact, what is to be summed in the sum over paths approach are not

96
5.2. The 2014 version

necessarily only the possible world-lines of a single particle, but more gener-
ally, all possible indistinguishable processes leading to the same experimental
outcome: indistinguishable, in the sense that no information can be retrieved
about which one of the processes has happened. An immediate consequence of
this generalization, which we will exploit in full in the 2015 sequence with the
introduction of the Hong-Ou-Mandel experiment, is the possibility of introduc-
ing the schematic representation of processes in terms of Feynman diagrams.

5.2.6 Massive particles

Observational tests of interference using massive objects prove that the quan-
tum behavior cannot be relegated to some special property of light. Besides
older setups using electron beams, such as the Davisson-Germer experiment
[378], single particle interference effects have been demonstrated for electrons
[379, 380], neutrons [381] and C60 molecules [382]. We base the extension of
the sum over paths formalism to the treatment of massive particles on the
following steps:
ˆ By reconsidering the photoelectric effect (leading to the photon energy E =
~ω) and summarizing the main conclusions that can be drawn from the
Compton scattering (~p = ~~k) we rewrite the photon phase φ = kx − ωt as
p E
φ= x− t (5.2)
~ ~
We postulate this formula to be valid for the phases carried by both massless
photons and massive particles alike.

ˆ From Eq.5.2, proceeding by way of analogy, we derive the expression for the
De Broglie wavelength, λ = hp .

ˆ In the case of a particle moving in a potential V (x) when the total energy
E is a constant of motion we write the more general form
Z
1 E
φ= p(x)dx − t (5.3)
~ ~
R Rp
where S0 =R pdx = 2m [E − V (x)]dx is the abbreviated action and
S = ~φ = L(x, t)dt is the Hamilton action [328, 383]. Again, we consider
the stationary case (fixed energy), in which the time dependent term in
Eq.5.3 plays no role in determining the phase difference between paths.

ˆ Eq.5.3 can be compared with the analogous expression for the phase of
monochromatic light entering an inhomogeneous medium with refractive
index n(x), i.e.
Z Z
φγ = k(x)dx − ωt = k0 n(x)dx − ωt (5.4)

97
 5. The design of our proposal

5.2.7 The limit of geometrical optics and classical me-

chanics: the correspondence principle
When the wavelength of the quantum object becomes much smaller than the
relevant length scales, the sum over paths approach reproduces the results of
a classical theory. For the photon, the dominant path in this limit is the one
predicted by the Fermat principle, thus retrieving the ray of geometrical op-
tics. The behavior of massive particles, on the other hand, approaches the
one predicted by the stationarity of the Hamilton action, leading to the cor-
respondence principle and explaining the emergence of classical mechanics as
an approximate theory. Simulations in which the wavelength (or the mass)
of the quantum object can be interactively varied are essential at this stage;
examples include light refraction at an interface, parabolic mirror reflection,
and diffraction of massive particles at a double slit (see Section 5.4).

5.2.8 Quantization in the sum over paths approach

In this part of the sequence we use the sum over paths representation of the
Green function developed in Section 4.4 to find energy levels and eigenfunctions
of quantum bound systems. In the first version of the sequence we only treated
simple examples, such as the particle restricted to a circular ring, which was
introduced by performing the analytical computation, and the infinite square
potential well, for which the allowed energy levels were also calculated ana-
lytically, while the shape of the resulting wave functions was shown by letting
students manipulate the corresponding simulation (see Section 5.4.7). In its
educational translation, the method presented in Section 4.6 consists in evalu-
ating the probability of detecting at position xD a particle emitted by a source
at xS by summing the amplitudes corresponding to all possible paths with a
fixed energy E, including those undergoing an arbitrary number of reflections
at the potential walls. The reflection coefficient at the barrier wall, which is
simply r = e−iπ is introduced through an analogy with optics. By taking the
square modulus of the resulting amplitude, the detection probability is com-
puted as a function of E. In the limit that the number N of paths considered
goes to infinity, only energies satisfying the condition that the vectors corre-
sponding to paths which differ by a full round trip in the well are in phase
produce a non-vanishing probability. In this way, the allowed energy levels, or
“spectra” for bound states can be defined, and analytical results can be proven
for simple systems.
The probability distribution for stationary states (i.e. |ψ(x)|2 ∝ G(x, x, E))
can be understood as the self-consistent distribution for the probability of find-
ing a particle, whose source is at x, at x again at some later time.
The sum over paths quantization of a particle on a ring also allows to introduce
the quantization of angular momentum, and to build a simple semiclassical
model of the atom, which is equivalent to the Bohr model in the De Broglie
interpretation. In fact, by fictitiously assuming that the electron is constrained

98
5.3. The 2015 sequence

to a circular “ring” around the nucleus, constructive interference between all

possible paths on the ring, including those performing an arbitrary number of
full turns, only happens in the case that the circumference length contains an
integer number of De Broglie wavelengths. Of course all the shortcomings and
inadequacies of the model have to be discussed here, but it must be stressed
that, in our approach, the Bohr model is not a point of departure, but more
a point of arrival. As students should by now have constructed a proper con-
ceptual structure for quantum mechanics, we believe that it should be much
easier for them, with respect to the initial stages of learning a new theory,
to understand and distinguish what is classical, and what is quantum, in the
model.

5.2.9 Comments on the structure of the 2014 sequence

The results of each one of the tests will be discussed in Chapter 6. Here,
however, we make some comments on the impressions we had, as the exper-
imentation was ongoing, about the effectiveness of our learning path in its
initial version.

ˆ First of all, the sequence appeared to be slightly unbalanced in the sense

that the initial part, presenting the Huygens principle and the transition
between the language of scalar optics and that of the sum over paths ap-
proach, had required too much time compared to its real importance in
sustaining students’ understanding. Part of the time invested on these in-
troductory subjects could maybe have used more proficuously to introduce
more advanced topics, or for experimental activities.

ˆ We were not entirely satisfied with the number and the variety of problems
and exercises which we had proposed or solved in class with students. In
fact, they were mostly specific examples about the sum over paths formal-
ism, which often bore little relation to exercises found in textbooks. Clearly,
a specific work was necessary on this point.

ˆ Some parts of the sequence had to be strengthened. In particular, more

experimental activities were needed, the difference between the classical
and quantum concepts of probability had to be more clearly focused, and
the presentation of the uncertainty principle had to be improved.

5.3 The 2015 sequence

Based on the detailed analysis of data (Section 6.1) and our subjective im-
pressions about the effectiveness of the sequence development in 2014 (Section
5.2.9) we operated relevant modifications to the sequence for the 2015 tests.
Except where we specify otherwise, the modifications were introduced in both
the test with ST and the one with high school students, which were carried

99
 5. The design of our proposal

out in an overlapping time span.

We devote a separate section (Sec. 5.5) to describe the wider setting of the high
school experimentation, where the sequence was expanded in several directions
to generate more complex [35] and stimulating learning environment.

5.3.1 Structure
The main innovations we operated in the 2015 version of the sequence can be
summarized as follows:

ˆ The first two hours of the sequence were devoted to experimental activities.
Besides the simple experiments on light interference and diffraction, the
activities concerned the measurement of h using LEDs of different color, and
the photoelectric effect (Sec. 5.3.2). In the course with ST, another activity
was proposed at the end of the course concerning the measurement of the
Rydberg constant using a home made spectrophotometer and the Tracker
[56] software (Sec. 5.3.10), while for high school students we proposed a
more restricted, qualitative version of the same activity which only involved
the observation of discrete spectra from gas discharge lamps.

ˆ The introductory part concerning the correspondence between Feynman’s

approach for the photon and the phasor formalism in wave optics was dras-
tically cut. The 2015 sequence effectively began, in the classroom part, with
the introduction of the photon and its properties. The analogy with wave
optics and the Huygens principle, however, was introduced in the context of
experimental activities on light diffraction and interference, and afterwards
frequently emerged during open discussions.

ˆ The presentation of the uncertainty principle was strengthened by discussing

different examples besides the case of diffraction from a slit of variable width.
The time-energy uncertainty principle was introduced as well, several ex-
ercised were proposed, and after the introduction of massive particles we
discussed the use of the principle for a qualitative explanation of the exis-
tence of a minimal energy for bound systems, and of the phenomenon of
tunneling [150].

ˆ By saving time in the initial part of the sequence we were able to leave more
space to applications and more advanced concepts. We introduced the time
independent description of open systems and tunneling through activities
with the simulations presented in Section 5.4.8. We also expanded the
presentation of atomic spectra and transitions, to connect with the proposed
experimental activities. Finally, in the course for ST only (mainly due to
time constraints in the experimentation in high school), we expanded on
an educational suggestion contained in QED, and introduced the concepts
of exchange of indistinguishable particles, the difference between bosons
and fermions, and the Pauli exclusion principle starting from the discussion

100
5.3. The 2015 sequence

of experimental evidence of the generalized Hong-Ou-Mandel experiment,

performed using both photons and electrons.

The general structure of the sequence adopted in 2015 is presented in Fig. 5.5.
In the following, we describe the the main steps of the sequence, concentrating
our attention on those which are new or significantly changed from the 2014
version.

Figure 5.5: Schematic representation of the teaching sequence for 2015. For
readability, discussed experimental results have been removed from the repre-
sentation. In this regard, the only addition is the Hong-Ou-Mandel experiment
discussed in Section 5.3.7.

5.3.2 Initial experiments

The current version of the sequence starts in the laboratory. Besides the simple
activities of Tracker analysis of interference and diffraction, discussed previ-
ously, we introduced two different experiments meant to measure the Planck
constant h. The first activity concerns the photoelectric effect; we do not de-
scribe it in detail as it is very common in university physics courses, and was
performed using conventional equipment of the educational laboratories at the
University of Pavia. In particular, we used the PASCOTM 9368 h/e apparatus
[384].

101
 5. The design of our proposal

In the second activity the Planck constant was measured using LEDs of dif-
ferent color [372, 373]. Since the theory underlying this second experiment is
perhaps not as well known as in the previous case, it may be useful to recapit-
ulate its essential lines.
When a LED is connected to a voltage source, it needs a minimum potential
difference Va to activate and start emitting light. This threshold value is related
to the frequency of the emitted light by an approximately linear relation

h φ
Va ' ν+ (5.5)
e e

Where e is the electron charge, and φ measures the energy loss for each elec-
tron inside the semiconductor’s p-n junction. Within the accuracy limits of the
experiment, φ can be considered constant for all LEDs. With this assumption,
the relationship between the frequency ν of light emitted by different color
LEDs (we used red, orange, green, and blue ones) and activation voltage Va is
linear.
The frequency of emitted light can be assumed as known from the character-
istics of the diode or, to make the experiment more complete and engaging,
can be directly measured using a simple spectrophotometer such as the one de-
scribed in section D. The activation voltage can be estimated simply by slowly
varying the potential difference and observing, in a dark environment, when
the LED starts emitting light. Some sources suggest instead to approximately
determine Va by regression from the linear characteristic of the LED for high
current, but we found this method to be less reliable since, at least in our case,
the Ohmic resistance for high currents was rather different from one diode to
another.
From Eq. 5.5 the value of h can be estimated by finding the slope of the least
square linear fit of the ν versus Va data. In our experience the values of the
Planck constant found through this procedure are within 10% of the accepted
value.

5.3.3 The photon concept

The 2015 sequence starts directly with the introduction of the photon and its
characteristics. The content of this step is essentially the same as the corre-
sponding one of 2014 (Sec.5.2.3), with very minor changes. For example, for
introducing the Grangier experiment on photon indivisibility we used the on-
line interactive representation of the experiment available at the site Quantum
Lab of the University of Erlangen [312]. Also, we included a more in depth
discussion of the Compton effect, which was only briefly touched in the first
version of the sequence.

102
5.3. The 2015 sequence

5.3.4 Feynman’s photon model

The content of this step is also approximately the same as the corresponding
one of the previous year (sec. 5.2.4). However, more exercises were discussed
in the classroom and assigned as individual work, for example those on the
determination of maxima and minima for the detection of photons in three
and four slits interference setups (See Appendix B).

5.3.5 Conceptual and foundational aspects, first part

The central core of our sequence (Sec. 5.2.5) had worked very well in 2014,
and remained mostly unchanged. In 2015 we added more exercises specifically
targeted at which way measurements and the calculation of probabilities. In
fact, we realized from open discussion with students that, when confronted
with the problem of finding the frequencies of outcomes from a two way inter-
ference experiment where an intermediate measurement was being performed,
considering the joint statistics (i.e. summing frequencies over all possible out-
comes of the which way measurement), a small number of them returned to
the familiar classical conception that “in this case, the frequencies must be the
same as in the case that no which way measurement is present”. Although,
as highlighted from answers to different questions, they may have formed a
correct mental model of the which way measurement selecting as possible only
those paths compatible with its outcome, they could not always relate such
model to a question in which the answer appears so trivial from a classical,
commonsense point of view. This may be seen as an example of how, in a
process of gradual conceptual change, two framework theories may coexist in
the mind of a learner, who may still, for a given question, select the answer
which is correct in the naı̈ve (or, in this case, classical) framework, which is
more easily accessible to him/her, even after having acquired the advanced
one [27]. Thus, in the perspective of guiding students in the development of
executive and procedural skills that could make easier for them to access the
new theoretical model, we created, discussed in class and assigned as home-
work exercises and problems exploring in detail the consequences of a which
way measurement on the distribution of probabilities for the outcomes of an
experiment (See Appendix B).
Still about the comparison between quantum and classical probability, an ob-
servation which some students, mainly those with a degree in mathematics in
the teacher training course, found illuminating is that actually the contradic-
tion between the classical and quantum ways of computing probabilities may
be seen as merely apparent. Indeed, the rule giving the probability for an event
E which can realize passing through two and only two alternative events A and
B as P (E) = P (E|A)+P (E|B) is only valid if A and B are mutually exclusive.
What quantum physics actually denies the validity of, and this can be seen
very clearly in the Feynman formulation, is not the rule itself, but rather the
hypothesis that A and B are mutually exclusive, i.e. that the proposition “a

103
 5. The design of our proposal

quantum object detected at E, when no which way measurement is performed,

must have passed through one and only one of the two slits A and B” can be
accepted as true.

5.3.6 Massive particles

The content of this rather short step is almost identical to its 2014 counterpart
(Sec. 5.2.6). The only thing we added is a video, showing the accumulation
of spots on a screen in a two slit experiment with one electron at a time [385]
which we introduced to visually demonstrate the perfect analogy between the
behaviour of photons and that of electrons with respect to Young’s experiment.

5.3.7 Conceptual and foundational aspects, second part

In 2015 a new phase for discussing conceptual and foundational aspects was
added, after the introduction of massive particles, by greatly expanding the
step which, in the initial version of the sequence, only included the discussion
of the correspondence principle. This was due to the fact that, reconsidering
the 2014 sequence, we found too limiting to only discuss conceptual aspects of
quantum theory using evidence from quantum optics, for two main reasons: we
wanted to introduce the Pauli exclusion principle starting from the generalized
Hong-Ou-Mandel (HOM) effect, and improve students’ understanding af the
uncertainty principle, presenting it also as an explanatory basis for typically
quantum phenomena such as the existence of a minimal energy for quantized
systems, and the tunnel effect.

5.3.7.1 The Hong-Ou-Mandel experiment

The original HOM experiment [296] demonstrates destructive interference be-
tween indistinguishable processes which only differ by the exchange of identical
photons. The setup consists of two “sources” (there may well be only one source
of photons in the usual meaning of the term, producing identical photons which
are then displaced, for example by means of optical fiber cables) emitting per-
fectly indistinguishable photons towards the two inputs of a beam splitter (Fig.
5.6 (a)). At the two outputs of the beam splitter are placed two detectors.
The possible outcomes of this experiments (Fig. 5.6(b,c,d)) are: (1) two pho-
tons are revealed at detector A; or (2) two photons are detected at B, or (3)
one photon is detected at A and one at B. This last outcome can be produced
through two indistinguishable processes, which differ by the exchange of the
two identical photons. Since the amplitudes of these two processes are out of
phase by π, because in one of the two the photon undergoes external reflection
from the beam splitter, the amplitudes of these two processes cancel with one
another4 . The end result is that the experimental outcome (3) is suppressed,
4
This result does not depend on the type of beam splitter used because of the general
property of beam splitters Eq. 5.9 deriving from the unitarity requirement.

104
5.3. The 2015 sequence

(a) (b)

(c) (d)

Figure 5.6: (a) Schematic representation of the Hong-Ou-Mandel experiment.

(b) Diagram for experimental result (1) in the text. (c) Diagram for experi-
mental result (2). (d) The two possible diagrams for experimental result (3).

and photons are always found at the same detector.

In the educational presentation of this experiment, already sketched by Feyn-
man in QED long before the experiment was actually performed, the following
steps are needed:
ˆ Introduce the concept of interference between indistinguishable processes, or
histories, between an initial and final state, involving more than one quan-
tum object. This can be done already starting from the ZWM experiment.
ˆ Introduce the rule for computing the amplitude for a process, as the product
(in the sense of complex numbers) of the amplitudes corresponding to each
individual particle involved in the process. As in our path (as in QED)
complex amplitudes are represented as vectors, the product of amplitude
must be explained as the operation of multiplying the moduli of the two
vectors, and adding up their phases.
ˆ Compute the amplitudes of the two processes leading to experimental out-
come (3), showing that they differ by a minus sign.

105
 5. The design of our proposal

In 2005 [359] the experiment was generalized by performing it with indistin-

guishable electrons. In this case, the beam splitter is replaced by a potential
barrier, especially tuned to the energy of electrons involved in the experiment
in such a way that each particle has a 50% probability of being reflected at the
barrier, and 50% of being transmitted. In this case, as predicted by the rules
of quantum physics, the experiment gave the opposite result: outcomes (1)
and (2) were suppressed, and the electrons were always found on at detector
A, one at detector B.
To interpret the experiment using a sum over histories approach, an additional
consideration must be made. The diagrams in Fig. 5.6 (b) and (c), repre-
senting experimental outcomes (1) and (2), can be thought as composed of
two sub-diagrams each, in which the two quantum objects arrive at points P
and P 0 at a vanishingly small distance from one another; and the situations
S1 → P , S2 → P 0 or S1 → P 0 , S2 → P are possible.
From this perspective, the experimental outcome can be explained by assum-
ing that, differently from the case of photons, a diagram obtained from another
by exchanging two identical electrons takes a minus sign.
The result can be generalized, allowing to divide quantum objects into two
separate categories 5 with respect to their behaviour in the generalized HOM
experiment:
ˆ Objects which in a HOM experiment behave like photons: if indistinguish-
able, they are always both found at the same detector. Amplitudes for
histories involving an exchange of two identical particles are computed with
the usual rules; these particles are called bosons.
ˆ Objects which in a HOM experiment behave like electrons: if indistinguish-
able, they are always found at different detectors. For such objects, ampli-
tudes must be computed in the following way: taken one arbitrary history as
a reference, all histories which differ from the reference one by an odd num-
ber of exchanges of identical particles, take a minus sign. These particles
are called fermions.
By generalizing even more, and considering the abstract case in which two
identical fermions, starting from different states A and B, could arrive both at
the same state C, and repeating the previous reasoning, we arrive to formu-
lating the Pauli exclusion principle: identical fermions can never occupy the
same quantum state.

5.3.7.2 Improving the presentation of the uncertainty principle

A second objective of introducing this new step was reinforcing student’s under-
standing of the uncertainty principle by presenting it not only as a statement of
5
Here the consideration can be added that if a multiplicative factor a is assigned to the
operation of exchanging the roles of two identical particles in a diagram, such factor can
only be either 1 or −1, because applying the operation twice gives back the original diagram,
and so it must be a2 = 1.

106
5.3. The 2015 sequence

impossibility, bu also as an explanatory principle of physical facts. We believe

that, by doing so, the possible mistaken conception of Heisenberg’s statement
as due to a disturbance in a measurement process can further be prevented. In
fact, by using the principle to explain objective physical phenomena in which
measurement is not even mentioned, it should become even more apparent that
the inequality ∆x∆px ≥ ~2 has to be interpreted as a fundamental constraint
on the quantum state. We discuss with students two examples:

ˆ The uncertainty principle proving the existence of a minimal allowed energy

for quantum bound systems.

ˆ The uncertainty principle providing a qualitative (and semi-quantitative)

explanation of quantum tunneling.

Concerning the first point, the argument can be found in several books (see
e.g. [386, 387]) and can also be applied to several potentials, including the
harmonic oscillator, Coulomb, and ramp [388]. It can be formulated in its
simplest form in the case of a particle confined in an infinite potential well
of width L. In this case, the uncertainty on the position of the particle is
∆x ' L2 . Then, according to the Heisenberg principle, ∆px ≥ L~ must hold.
From statistics, the following general relation holds:

(∆px )2 = p2x − hpx i2


(5.6)

where the brackets denote ensemble average. Since hpx i = 0 by symmetry

considerations, (∆px )2 = hp2x i and we obtain the following inequality for the
(averaged) energy of a state:

hp2x i ~2
hEi = ≥ (5.7)
2m 2mL2
i.e. the energy of the quantum object must be greater of a minimal value,
which (coincidentally) in this case is the exact value of the bound state energy.
The argument for the second point can also be found in several sources, with
variations. In particular, we used the discussion by Lévy-Leblond [267] which
we do not reproduce here. The qualitative discussion of the tunnel effect using
the uncertainty principle leads to the next step of the sequence, where the
same phenomenon is treated exactly.

5.3.8 Open systems and the tunnel effect

Tunneling is one of the signature phenomena of quantum physics, and can be
used to explain other phenomena, such as alpha radioactive decay [389], and
important application such as the electron tunneling microscope. In the liter-
ature, tunneling is often reported as problematic for students in introductory
courses (see Sec. 3.1.1.4); for example, they often misinterpret tunneling as a
dissipative process in which the particle loses energy at the barrier, a problem

107
 5. The design of our proposal

which has been connected to inaccurate representations of the wave function,

and interference with classical concepts [113, 201]. A consistent introduction
to time independent tunneling using the sum over paths approach for the quan-
tum object was one of the principal objectives of the sequence revision.
As we have demonstrated in Section 4.6 of this dissertation, the problem of
tunneling from a rectangular potential barrier is exactly solvable using the sum
over paths representation of the Green function; furthermore, the usual formu-
lation of this problem is essentially already expressed in the source-to-detector
philosophy, which makes the educational translation entirely straightforward.
A source is placed on one side (usually, to the left) of the potential barrier, and
the probability must be computed that it reaches a detector on the right side.
This probability, which apart from a constant factor corresponds to the Green
function G(xS , xD , E), can be obtained by summing the amplitudes from all
paths performing an arbitrary number of reflections inside the barrier, taking
into account transmission and reflection coefficients, exactly like in the case of
the reflection of light from a film of dielectric material 6 . Notice that these
paths can all (except at most one) be considered non classical, because the
classical particle never undergoes internal reflections.
The amplitudes for the first two or three paths at a given energy can be com-
puted by hand, providing insight on the functioning of the method, and giving
a good approximation to the exact result. Considering only one path gives a
very rough estimate of the transmission factor, which ignores resonances, and
corresponds to using the WKB approximation. For a full solution of the prob-
lem, the simulation shown in Sec. 5.4.8 can be used, which allows to increase
at will the number of paths considered.
It is important to clarify to students that the entire process happens at fixed
energy, so this physical quantity is exactly the same for the quantum object
before or after the barrier. The transmission probability, however, does depend
on energy 7 .
We then turn to the analysis of a semi-open, resonant system, in which the
detection probability depends very strongly on energy, showing sharp reso-
nance features. The system, in the optical case, consists of a source emitting
monochromatic photons towards a detector, with two successive 50/50 beam
splitters interposed along the path. In the case of massive particle, as already
seen in the case of the generalized Hong-Ou-Mandel effect (Sec. 5.3.7) such
role can be played by thin potential barriers of variable height, especially tuned
in such a way that for any given energy the square modulus of the transmission
coefficient is always |t|2 = T = 0.5. By varying the energy E of the incoming
quantum object, one observes that the detection amplitude goes from a very
6
As for the r = e−iπ factor in the previous version of the sequence, the general form of
the transmission and reflection coefficients Eq. 4.42 are introduced through an analogy with
optics.
7
This implies that, in the time dependent case, the mean energy of the transmitted “wave
packet” is in general different from that of the incoming one. However, the energy is not
necessarily lower; in fact it is typically slightly higher [390].

108
5.3. The 2015 sequence

low minimum, to very sharp maxima, which, as usual, correspond to the case
in which all amplitudes associated to two paths differing of a full back and
forth reflection in the two barrier system are in phase. This system does not
yet have a discrete set of energy levels; however, it strongly selects some values
of energy, for which the probability of detection is much higher.

5.3.9 Bound systems and quantization rules

The semi-open resonant system introduced in the previous section introduces
seamlessly the transition to the treatment of bound systems. In the 2015 ver-
sion, this step of the sequence has been substantially reinforced. First of all,
after the “particle in a box” case, the square potential well with finite walls
height was discussed also. This was natural since, after having introduced in
the previous section the example of square barrier tunneling, this case could
be treated in a rather straightforward manner, without introducing new con-
ceptual tools. More importantly, we improved the treatment of the atomic
model and transitions, with the dual objective of not letting students stick to
the Bohr model, and to give them the tools to correctly approach the usual
textbook exercises on atomic transitions.
To reach these objectives, our presentation was carefully rethought on two
main points:

ˆ When introducing the simple model of the hydrogen atom, we payed more
attention to make clear to students that such model, based on the assump-
tion that the electron paths to be considered are only two dimensional, clas-
sical, circular orbits around the nucleus is completely unrealistic for more
than one reason. Indeed, it can even be seen as surprising that the model, for
the hydrogen atom, gives reasonable results. A consistent treatment would
have to take into account, as we in fact did for the simple one dimensional
systems we had previously solved, both classical and non classical paths; and
this should have been done in a full three dimensional space 8 . However,
the general principle still stands, that the allowed energy level represent a
condition of constructive interference between the infinitely many paths at
a given energy. The correct result for the Hydrogen atom, which can be
introduced qualitatively to students, is a stationary three-dimensional wave
function for each energy level, whose square modulus gives the probability
distribution of finding the electron at a certain point in space for each one
of the allowed energy values.

ˆ In order to prevent the graphical representation of the Bohr model from

consolidating into students’ minds [23], immediately after its introduction
8
The hydrogen atom problem was exactly solved using the path integral approach by
Duru and Kleinert in 1982 [391]. A sophisticated, but still semiclassical sum over paths
approximation using only all classical periodic orbits is known as Gutzwiller’s trace formula
[342, 392].

109
 5. The design of our proposal

we substituted it with the energy level representation, which describes the

atom’s physical behavior but not its physical properties, as suggested for
example in [168, 112]. Afterwards, we consistently used this representation
in all tests and exercises proposed to students (See Appendix B).

5.3.10 Measurement of the Rydberg constant

As the theoretical part of our sequence ends with the discussion of energy lev-
els, atomic spectra and transitions, it was natural to conclude our journey into
quantum physics with the observation of atomic spectra, and the measure-
ment of the Rydberg constant. This activity is also widely used in educational
settings (see e.g. Ref. [374]); in this case our main innovation consisted in
the use of a spectrophotometer realized with low cost materials. The realiza-
tion and calibration of such spectrophotometer is described in Appendix D.
Schematically, the activity develops according to the following steps:

(a)

(b)

Figure 5.7: (a) Spectrum lines (red, cyan and blue) clearly visible in the pho-
tos obtained using the low cost spectrometers. (B) Graphs of intensity versus
wavelength for the hydrogen lamp spectrum obtained with two different spec-
trometers (see Appendix D for details).

1. Using a basic apparatus (i.e. the “body” of the spectrophotometer) made

from black cardboard, a piece of transmission or reflection grating, and a
photo camera, students acquire images of the spectrum of a common mer-
cury vapor discharge lamp used for domestic illumination, and the spectrum
of an hydrogen gas lamp.

110
5.3. The 2015 sequence

2. Working first on the photos of the spectrum of the mercury vapor lamp, and
using Tracker to determine the peaks of the spectrum, students calibrate the
wavelength measurement of the spectrophotometer, taking as a reference
two lines which are invariably present in mercury gas lamps, i.e. the 436
nm (blue) and 536 nm (green) lines.

3. Students can now determine the wavelengths of the peaks of the spectrum
of the hydrogen gas lines, which are the lines of the Balmer series. In
particular, as it is usual for most, even professional spectrometers, the red
(ni = 3), cyan (ni = 4) and blue, (ni = 5) are usually visible (see Fig.
5.7), as the fourth possible line is very close to the violet edge of the visible
spectrum.

4. Using the formula

!
1 1 1
=R 2
− 2 with ni > nf (5.8)
λ nf ni

Where R is the Rydberg constant, and for the Balmer series nf = 2, they
can determine the value of R, either by evaluating the slope of a linear fit
of λ1 versus n12 , or simply by computing a separate value of R for each of
i
the three spectral lines and taking an average.
Using this method, students can typically evaluate the Rydberg constant
to less than 1% difference from its accepted value R = 1.097 × 107 m−1 [59]

5.3.11 Comments on the 2015 sequence

Leaving a complete evaluation of results to Chapter 6, our subjective impres-
sions on the refinements operated in the 2015 version of the sequence were
extremely positive. The learning path was more equilibrated, with a more
proportionate repartition of time between experimental activities, theory, use
of simulations and exercises. The experimental activities were satisfying for
students, and exercises were more focused on points in which the learning pro-
cess could encounter significant obstacles. The number of subjects covered
was significantly expanded, without impairing the formation by students of
coherent and stable mental models and “pictures” on the central conceptual
and foundational themes of quantum physics, which was the main focus of the
2014 sequence.
Several of the changes introduced went in the direction of improving the an-
swer to the research questions formulated in Section 1.2; for example, the
introduction of the Hong-Ou-Mandel experiment, which is one of the founding
results of the field of quantum optics (Q1, Q5); the design improvement on the
presentation of the uncertainty principle (Q4); the addition of new low cost
experiment (Q7), simulations (Q6) and exercises specifically targeted to deal
with students’ difficulties (Q3).

111
 5. The design of our proposal

Some important themes are still not treated in the course we designed, includ-
ing, to name only the two most important ones, the behaviour of objects with
spin (except for the Pauli exclusion principle) and time evolution. Projects
for future expansions of the sequence, as well as possible refinements on the
currently covered material, are discussed in detail in Sec. 7.2.

5.4 Details of GeoGebra simulations

Simulations traditionally employed with the sum over path approach refer
to diffraction and interference at slits experiments. We tried to enlarge the
panorama by exploring other experimental situations such as the Lloyd mirror,
demonstrating a π phase loss for “external” reflection, and the Mach-Zehnder
interferometer, for which we designed a rather complex simulation including
removable arms. We tried to render simulations as interactive and visually
engaging as possible and to make them as close as possible to real experimental
situations.

5.4.1 Interference and diffraction

Figs. 5.8 (referring to diffraction.ggb) and 5.9 ( grating.ggb) show the general
form of most of our simulations: two separate graphical views are paired in
adjacent windows, while the algebra view is usually left hidden, although it
can be brought up if necessary. The left graphical window is used for the
geometrical setup of the experiment, for interactive sliders and boxes, and
visualization of the final probability distribution.
The right graphics window shows the sum of phasors corresponding to the
paths arriving to the detector point (which can be chosen through a slider).
We found this general structure to be easy to understand and manipulate for
users, and to provide a good understanding of the functioning of the sum over
paths algorithm.
Fig.5.8 shows diffraction of photons from a single slit of variable width. This
case plays a very important role in our educational path, as it is connected
to the introduction of the uncertainty principle. We found that, for students,
observing how the phasors add up in this case, and in particular how they ap-
proximately form a circle at the first minimum, is an useful complement to the
explanation found on textbooks, which they often find difficult to understand,
being based on symmetry considerations.
Fig.5.9 shows diffraction of light from a grating. This simulation can be used
for an intuitive introduction to selective light diffraction, a phenomenon which
we explore in experimental activities such as the observation of spectra pro-
duced by gas discharge lamps, and the measurement of the Rydberg constant
(Sec. 5.3.10). In this case we found useful to restrict the possible values of the
wavelength in the simulation to those of visible spectrum of light. The proba-
bility distribution for a photon on the screen appears colored correspondingly

112
5.4. Details of GeoGebra simulations

Figure 5.8: Single slit diffraction. In the left window the final probability
distribution is represented, and (if desired, through checkboxes) the real and
imaginary parts of the amplitude can be visualized, drawn in a lighter stroke.
In the right window, the construction of the resultant arrow at the detector
point, with the phasors corresponding to individual possible paths (small red
arrows) and the resulting arrow (black, in a heavier brush stroke). The width
of the slit can be varied through the parameter d. This simulation is used to
provide a qualitative understanding of the uncertainty relation.

to the selected wavelength. Thus, the secondary peaks spread out and change
color as the wavelength increases, providing a visually effective representation
of selective light diffraction.

Figure 5.9: Sum over paths simulation of a diffraction grating. Parameters

that can be varied include the grating pitch d, the light wavelength (values
allowed here correspond to those in the visible spectrum of light, in nm), the
source-detector distance D.

113
 5. The design of our proposal

5.4.2 Double slit interference

Young’s two slit experiment is one of the main cornerstones of the educational
path. It is used to introduce the method of Feynman paths, starting from the
observation of the effect of performing the experiment with one photon at a
time. Then, we return to it to exemplify several other concepts: the quantum
rule for probabilities, the effect of acquired information on the possible paths,
or “wave particle duality”, and the quantum behaviour of massive particles
such as electrons.
. Initially, we start with a perfect correspondence with the usual treatment
of the phenomenon in wave optics, and use a simulation with pointlike slits
(Fig. 5.10 (a) and simulation (twoslits-1.ggb)). Since the most important goal
here is to perform the transition between the language of wave optics to that of
quantum physics, learning think about the probability of detecting the photon,
rather than the intensity of light, the simulations used need not be unnecessar-
ily complicated; however, we found useful to include a checkbox to take into
account the 1r factor in the photon propagator, since the probability density
which results from ignoring it (which is not normalizable, and so unphysical)
did not appear realistic to students.
Later in the path, when we return to the two slit experiment for highlighting
more advanced concepts, we use a simulation with finite slit width (Fig. 5.10
(b) and simulation (twoslits-2.ggb). In this case we emulate the effect of an
ideal, “non interacting” observer using checkboxes which allow to exclude from
the sum of phasors the paths corresponding to the slit at which the photon
has not been observed. In this way, the simulation can be used to show the
relation between the classical and quantum concepts of probability, and also
the role of information acquired on a quantum system.

5.4.3 Refraction at an interface

In this case we proposed two different simulations (Fig.5.11). The first one
(refraction-1.ggb) is a re-implementation of a classical example already pro-
duced by several authors [5, 308, 336]. The experimental situation represented
is an isotropic light source near an interface between two media, with a detec-
tor placed beyond the interface. The shape of the “Cornu spiral” which appears
in the right graphic window in Fig.5.11(a) shows that the paths nearby to the
one of minimum time give a larger contribution to the final amplitude. Also,
the minimum time path becomes more and more dominant, as the wavelength
decreases with respect to the source-detector distance. Paths which are very
far from the ray of geometrical optics give essentially no contribution, since
they go round in “curls” at the ends of the spiral. We also produced a Ge-
oGebra file (refraction-2.ggb) representing the phenomenon of refraction at an
interface as seen in a different experimental setup, more similar to those used
in real experiments. In this case the situation represented is the pointing of a
“collimated beam” towards the interface with a certain variable incidence an-

114
5.4. Details of GeoGebra simulations

(a)

(b)

Figure 5.10: Sum over paths simulations of two slit interference. (a) Pointlike
slits. It is possible to take into account, or ignore, the effect of path length on
the amplitudes associated to paths. (b) Slits of finite width. In this case the
effect of a perfect “non interacting” observer at one of the slits can be emulated

gle α. The resulting probability distribution found at some distance inside the
material has a peak at the position given by Snell’s law. Since 9 discretization
points are used for the source, and 5 for the interface, 45 paths are computed
for each detector point (in this case, the phasor sum is not visualized). This
example is meant to show relatively advanced students how somewhat realistic
situations can be modeled using the sum over paths approach.

115
 5. The design of our proposal

(a)

(b)

Figure 5.11: Refraction of a photon at the interface between two materials

with different refraction indices n1 and n2 . (a) Traditional way of representing
the phenomenon, with paths nearer to the ray of geometrical optics shown in
a different shade (shifting towards red in the online version). (b) A different
experimental arrangement for the same phenomenon, showing the bending of
a collimated ray directed towards an interface.

5.4.4 Lloyd mirror and phase loss upon reflection

The explanation of some interferometry experiments in the sum over paths

perspective needs a detailed analysis of the consequences of reflection at an

116
5.4. Details of GeoGebra simulations

9
interface. For example, a general property of beam splitters is the relation

δ1 + δ2 = π (5.9)

where δ1 and δ2 are the phase shifts between transmitted and reflected rays at
the two inputs [393]. The peculiar features of the kind of Mach-Zehnder in-
terferometer which we represent in our simulation depend on the use of beam
splitters with δ1 = π and δ2 = 0, that is, in which the reflected ray has a
π phase shift with respect to the transmitted ray on one of the inputs, but
no phase change on the second one [394]. This can be obtained by using the
most common form of beam splitter, i.e. a cube made from two triangular
glass prisms which are glued together at their base using polyester, epoxy, or
urethane-based adhesives. In the simulation we use a pictorial representation
of the beam splitter as a simple glass prism, as this representation helps stu-
dent in distinguishing external reflections (which cause a π phase loss) from
internal ones (which do not produce any phase shift).
As discussed in Section 4.4, applying to phasors associated with Feynman
paths the appropriate phase shift upon reflection is essential for obtaining the
correct energy levels for bound systems. For these reasons we discuss the phase
change associated with reflection early in our sequence, using the Lloyd mirror
experiment and simulation. Other reconstructions neglect to point out the
phase change at reflection because they do not consider experiments where
interference between reflected and direct paths takes place. The Lloyd mir-
ror setup (Fig. 5.12, file lloydmirror.ggb) consists in a light source emitting
photons, which can reach a detector (screen) either through a direct path, or
after being reflected by a mirror placed orthogonally to a the screen. The

Figure 5.12: Simulation of the Lloyd mirror.

discussion of this experiment is simple: the interference pattern between the

9
This property is a consequence of the requirement that the matrix operator representing
the beam splitter is unitary

117
 5. The design of our proposal

two possible photon paths has a minimum at the origin, showing that the di-
rect and reflected paths (which, since the point at the beginning on the screen
coincides with the point at the end of the mirror, have identical lengths) are
in phase opposition. This proves that the phase associated to a path reflected
on a mirror surface must be shifted by π. The setup can also be discussed in
terms of a “virtual source” placed inside the mirror, which is just the virtual
image of the real source. Using this strategy of analysis, the results from the
Lloyd mirror can be understood in terms of interference between two coherent
sources. With respect to the case of two slit interference, however, the dark
and bright fringes are reversed, because of the phase inversion of the paths
coming from the virtual source (the reflected paths).

5.4.5 Mach-Zehnder interferometer

The Mach-Zehnder interferometer (Fig.5.13, file machzehnder.ggb) is useful
both as a conceptual tool [137, 286, 288] , and as a training ground for per-
forming actual calculations. In its base form, the experiment is used with the
two arms having the same optical path length, so no phase shift is introduced
between the two photon paths. In this case, because of the π phase loss intro-
duced by beam splitters for external reflection, but not for internal reflection,
one of the detectors (Detector 2 in Fig.5.13) has zero probability of detecting
the photon, while the other detects the photon with certainty.
One of the main points of interest in the experiment is to compare this result
to what happens when either one of the two arms has been blocked: in this
case, detectors A and B have the same probability P = 12 of detecting the
photon, since no interference happens. Thus one can conclude that results are
incompatible with the hypothesis that, in the experiment with the full setup,
the photon has gone through only one of the two possible paths. Although it
is certainly possible to introduce the same idea using the two slit experiment,
we found that reducing the outcome possibilities to only two detectors, rather
than a continuous screen, has definite educational advantages.
The same example also quite effectively illustrates the differences between the
classical and quantum ways of computing probability.
In the simulation we built, checkboxes allow to compare the two situations,
highlighting the impossibility of assigning a definite path to the photon.
Also, the optical path of one of the arms can be varied through the insertion of
a dielectric film (reflection at the interface is neglected in this case) of variable
width. This offers the possibility for students to actually compute probabilities
of detection, in a situation which is slightly different from the usual case of
two slit interference.
Exercises on the Mach-Zehnder interferometer and activities involving this sim-
ulation are very useful to students for two reasons: first of all, the setup only
involves paths having the same physical length; so, in order to compute the
effects of interference, students only need to count the “true” (producing a π
phase loss) reflections along the path, and the resulting vectors at the detector

118
5.4. Details of GeoGebra simulations

Figure 5.13: Simulation of the Mach-Zehnder interferometer. Parameters to be

freely varied include the width of a dielectric film which varies the optical path
length at one of the arms of the interferometers (which can also be removed by
setting its width to zero), and the photon wavelength. Checkboxes “Remove1”
and “Remove2” reduce the interferometer setup to only one of the two arms,
in both cases leading to equal probabilities at detectors.

will either be in phase or in phase opposition, making calculations extremely

easy. The relatively more complicated case in which a phase shift is introduced
in one of the arms of the interferometer is also very meaningful because it can
be used to introduce to students the concept of normalization of amplitudes.
Indeed, up to this point students have always been told that the probability of
detecting a photon at a certain detector is proportional to the squared modu-
lus of the resulting amplitude, and indeed, in exercises concerning interference
setups with multiple slits which are proposed to students before this point of
the path, it was never required to compute absolute probabilities of detection,
but only ratios between probabilities. In the Mach-Zehnder setup the photon
basically can only arrive either at detector A or at detector B; so, in this case,
students can actually compute probabilities, by assuming that the probabil-
ity that the photon arrives at A or at B can be set to unity. Starting from
this example, the general concept that probabilities (and amplitudes) must ba
normalized in such a way that the total probability of finding the quantum
objects at any one of the places where it may go, must be set to one.
Finally, we note, although for the present moment this possibility has not
been exploited in our sequence, that the Mach-Zehnder interferometer, when
described in terms of two component state vectors and matrices correspond-
ing to its two inputs and outputs, is formally identical to the Stern-Gerlach
apparatus for spin 21 particles. Thus, the apparatus can in principle be used
as a Rosetta stone [395] connecting the Feynman and Dirac formalisms and
allowing to proceed with the treatment of spin and light polarization.

119
 5. The design of our proposal

5.4.6 The limits of geometrical optics and classical me-

chanics
We present simulations aimed to show students the emergence of classical the-
ories in the short wavelength limit. The first one (Fig.5.14, file parabolic.ggb)
represents the parabolic mirror. It is well known that the Fermat principle ap-
plied to the case of a parabolic mirror leads to prove that all rays coming from
a very distant source (ideally at infinite distance) perpendicularly to the direc-
trix of the parabola are reflected towards its focus. Using the sum over paths
approach with a mirror surface, whose focal distance is comparable with the
wavelength of incident light, leads to a Gaussian-like probability distribution
of the photon being detected along the axis of the parabola. This probability
distribution has a maximum in the focus, and becomes more and more nar-
rowly concentrated around the peak as the photon wavelength decreases and
becomes negligibly short with respect to the relevant length scale (i.e. the focal
distance of the parabola). Simulations of simpler cases, such as reflection from
a plane mirror reflection.ggb) or free propagation of light in vacuum freeprop-
agation.ggb) in which, as the wavelenghth is varied, the paths corresponding
to the central part of the Cornu spiral (i.e. those which contribute more to the
final amplitude) are highlighted can also serve to explai the emergence of ray
optics from Feynman’s model [153].
The second simulation (Fig.5.15, file twoslits-3.ggb) represents a double slit
experiment with finite length slits and massive particles. In this case, the
wavelength corresponds of course to the De Broglie wavelength λ = hp , and
is inversely proportional to the particle mass. In the simulation the particle
mass is indeed the most important variable parameter; and the result of the
two slit interference is a well defined interference figure for small mass values,
which transforms as the mass increases into the classical expected result of two
“heaps” of particles, one behind each of the slits. Of course in this case also the
above discussion relating the wavelength of the particle to the relevant length
scale applies.

5.4.7 Simulations of time independent bound systems

In our educational path we used the “particle in a box” model as a first ex-
ample of the treatment of bound systems in the sum over paths approach.
We produced simulations for both the cases of the infinite (Fig. 5.16, files
infinitewell-1.ggb, infinitewell-2.ggb) and finite potential well (not shown). The
underlying mathematical algorithm is the sum over paths representation of the
Green function, which was developed in Section 4.4.
We use the “particle in a box” model as the paradigmatic example for the
treatment of bound systems in the sum over paths approach (Fig. 5.16). A
quantum object is confined in a square potential well with infinite depth and
width AB = 2a . We take one of the points inside the hole, which can be
chosen through sliders, to be the Source (S), and another one to be the Detec-

120
5.4. Details of GeoGebra simulations

Figure 5.14: Simulation of a parabolic mirror. In the case shown, the detector
point is near to, but not coincident with, the parabola focus. In the simulation,
besides the detector position, the photon wavelength can be varied to demon-
strate convergence to a Dirac delta-like distribution for small wavelengths.

(a) (b)

Figure 5.15: Simulation of two slits interference with massive particles: varying
the particle mass, one obtains either quantum interference (a), for small values
of the mass or classical particle heaps (b) for high mass values. Some of
the small ripple effects in the right figure are actually spurious – a numerical
artifact due to the issue that when the wavelength considered becomes too
small, interference effects start appearing between the points by which the
slits have been divided. This issue is always present in the simulations, and
prevents the possibility of reducing the wavelength too much; but of course it
is an effect which would in principle disappear in the continuum limit

tor (D). Four basic possible routes exist from the source to the detector: the

121
 5. The design of our proposal

direct S → D path, the S → A → D and S → B → D reflected paths, and

the S → A → B → D comprising two reflections.
Theoretically, all the paths which can be constructed by adding to any of the
above an arbitrary number of full back and forth routes should be considered;
in this simulation, we only add up N = 5 derived paths for each one of the
four basic types. The phasor corresponding to each path is computed by the
usual rules, with each reflection contributing a π phase loss. In Fig. 5.16 the
main characteristics of the sum over paths approach to bound systems can be
observed: if the energy corresponds to one of the potential eigenvalues, the
paths differing for an integer number of back and forth trips in the hole are
in phase, and the corresponding resultant amplitude has a peak 10 . For all
other values, the amplitude in the limit N → ∞ will vanish. Parameters that
can be varied include the well half-width a, the positions of the source and
detector, and the energy of the quantum object. The simulation concerning
the well with walls of finite height also allows to vary the potential height V ,
and the number N of paths considered for each of the basic families. As the
parameter N is increased, the spectrum of allowed energy levels visible in Fig.
5.16 becomes more sharply defined.
Inside the well, the simulation also plots the stationary probability density
|ψ(x)|2 = G(x, x, E). This quantity is computed and plotted by the simula-
tion in all cases, but for E 6= En it is almost vanishing everywhere because of
destructive interference (as usual it completely vanishes in the limit N → ∞),
while it appears when the energy of the particle corresponds to one of the
allowed energy levels (or in the finite N case, when it is very close to it).

5.4.8 Simulations of time independent open systems

We produced two examples of simulations of open systems, which are shown
in Fig. 5.17. The relevance of the problems represented in the simulations,
i.e. double barrier resonant tunneling (doublebarrier.ggb) and square barrier
tunneling (squaretunneling.ggb), has been discussed in Sec. 5.3.8. The two
simulations are similar in structure. In both cases, the left window contains
the geometrical setting of the problem, the adjustable parameters including
the number of paths form source to detector to consider, the graphical repre-
sentation of the sum of amplitudes at the detector, and a yellow bar whose
height is proportional to the energy of the particle. The right window contains
the computed probability as a function of energy, which is graphed by a pink-
colored point which moves and is tracked as the energy changes. In Fig. 5.3.8
(a) resonances, which constitute the main point of presenting this example,
are clearly visible. In Fig. 5.3.8 (b) the computed probability is compared to
the exact result. The main reason for performing such comparison is to visu-
ally see the progressive effect of adding more paths to the single paths rough
10
Since G(xS , xD , E) ∝ ψ(xS )ψ ∗ (xD ), this is true provided the chosen source position xS
is not a node of ψ(xS )

122
5.4. Details of GeoGebra simulations

(a)

(b)

Figure 5.16: Sum over paths simulation for the infinite square well. (a) if the
energy E of the particle does not correspond to an eigenvalue, the phasors
(corresponding to single path propagators) sum destructively, producing a re-
sultant amplitude which goes to zero in the limit that the number of paths
considered goes to infinity. (b) If the energy corresponds to an eigenvalue En ,
phasors add up constructively. Notice that, in this case, the sum of phasors
is a real number (the vector sum is superposed to the x axis.) Note that
for certain choices of the source or detector positions not all the eigenvalues
will be visible as peaks; in particular, those corresponding to wave functions
possessing a node at the source or detector position will disappear.

approximation (which corresponds to the WKB approximation). If more than

three internal reflections are considered, the computed energy is practically
identical to the exact one. In the tunneling simulation the mass of the particle
can also be varied; this possibility is not present in the simulation of the reso-
nant system, which is best understood in the optical case. On the contrary, in
the simulation of Fig. 5.3.8 (a) the transmission / reflection probabilities of th
two beam splitters can be varied, making them different from the basic 50/50
case.

123
 5. The design of our proposal

(a)

(b)

Figure 5.17: Simulations of open systems. (a) Resonant tunneling of a photon

from two successive beam splitters. (b) Square barrier tunneling

5.5 Rethinking the sequence for the 2015 test

in high school
In 2015 we tested the sequence directly in high school, with a last year class of
Liceo Scientifico Enrico Fermi in Genova. For this experimentation, we decided
to re-think our learning path, with two main objectives in mind.

ˆ From an educational-pedagogical point of view, we felt that, for the high

school experimentation, it would have been appropriate to pursue a wider
goal than student’s acquisition of the subject matter, trying at the same time
to make the disciplinary content relevant for them, and make it become part

124
5.5. Rethinking the sequence for the 2015 test in high school

of their cultural and personal identity.

ˆ From the point of view of physics education research, we felt that it would
have been interesting to compare, with respect to the goal of stimulating
students’ appropriation, our results with those of group of the University
of Bologna, who were at the time starting an experimentation of a learning
path that used a very different approach to the disciplinary content, but
had similar educational goals.

In order to meet these objectives, we had to significantly expand some aspects

of our sequence, in particular those related to epistemological discussion, and
to introduce new themes and activities. In particular, according to the gen-
eral theoretical perspective, which we discussed in Sec. 2.2, of the approach
of Levrini et al. [1, 34, 35, 396], a “properly complex learning environment”
intended to favor appropriation needs to respect the three general criteria of
“multi-perspectiveness”, “multi-dimensionality”, and “longitudinality”. Among
these three criteria, longitudinality is perhaps the most widely recognized as a
requirement for educational reconstructions of physical content (see e.g. Ref.
[71, 36, 397]). This criterion had been explicitly taken into consideration in
the design of our proposal.
Multi dimensionality, in the sense of making the epistemological structure of
the discipline visible, [75] had also been a primary concern in our design, as the
choice of the Feynman approach was, from the start, connected to the possibil-
ity of verbalizing a conceptual structure that may remain hidden or unspoken
in other formulations. The multidimensional character, however, had to be
made more explicit to students, and metacognitive reflection had to be stimu-
lated.
Finally, although our sequence contained some elements of multi-perspectiveness,
expanding the learning environment under this respect was certainly necessary.

5.5.1 Longitudinality
The criterion of longitudinality applied to an educational reconstruction means
that it should properly take into account that “learning physics is a continuous
process of widening, refining and revising already acquired knowledge” [34]. In
this perspective, the learning path should frequently include occasions in which
the new model is compared to theories previously studied, and the coherence
of the global picture is discussed. We briefly summarize the ways in which
longitudinality emerges from our sequence :

ˆ In the first part of the sequence, Feynman’s approach to the photon is

frequently compared with the wave model of light, explaining how the two
views provide different perspectives on interference and diffraction of light,
but only the former is able to explain those phenomena when only one
photon at a time is involved.

125
 5. The design of our proposal

ˆ The emergence of classical mechanics and geometrical optics from Feyn-

man’s model in the limit of small wavelengths is discussed in depth.

ˆ The quantum and classical concepts of probability, both in simple phenom-

ena and in the context of which way measurements are frequently juxtaposed
and compared (Sec. 5.3.5). Also, again in the classical limit, the two ways
of computing probabilities are reconciled.

In the high school context we certainly tried to highlight the longitudinal char-
acter of the sequence; these can be considered normal adjustments of the
“cloud” part of the sequence [40] which are necessary in transferring it to a
different setting, and we do not describe them in detail; for example, we ex-
panded the critical reflection on the photoelectric effect as demonstrating the
inadequacies of Maxwell’s wave model of light. One activity stimulating lon-
gitudinal reflection which we introduced anew in the high school test was the
first part of the questionnaire described in the next section.

5.5.2 Multi-dimensionality
The criterion of multi-dimensionality can be understood as the requirement
that, in the educational path, the content is analyzed and discussed “according
to different dimensions involved in physics, i.e. for their conceptual, experi-
mental and formal implications, but also for their philosophical-epistemological
peculiarities.” [34]. Apart from the evident connections with the experimental
dimension, the core nucleus of multidimensionality in our sequence was the idea
of making the conceptual and epistemological structure of quantum physics vis-
ible and verbalizable, using Feynman’s formulation. This basic concept had to
be made transparent and explicit to students, while at the same time stimulat-
ing meta-cognitive reflection. Based on the examples in the literature [34, 396]
we expanded the above concept into a wider and more general epistemologi-
cal discourse on physical theories running as an underlying thread during the
development of the sequence. Precisely, the theme concerned the possibility
of providing an intuitive representation of physical theories by means of im-
ages, analogies, or mathematical models, and was operationally introduced by
a questionnaire in two parts, which is reported in Appendix A.

ˆ The first part concentrated on three physical phenomena or principles which

students had previously encountered: the Joule effect, the second principle
of thermodynamics, and the magnetic force on a current carrying wire.
For each of these subjects, students were asked if they could recall (a)
Formulas or elements of the mathematical model. (b) An intuitive account
of the phenomenon or principle. (c) A visual, pictorial representation. This
part was meant to introduce a meta-cognitive perspective, leading each
student to reflect on his own way of learning and understanding physics.
In this way, students were encouraged to collect elements for providing an
organized answer to questions of the second part, and in particular to the

126
5.5. Rethinking the sequence for the 2015 test in high school

last one, which required them to take a personal position. This part of the
questionnaire can also be connected to the criterion of longitudinality, as it
touches phenomena connected to different areas of physics.
ˆ The second part of the questionnaire included quotations by physicists of
the past, such as Maxwell [398], Hertz [399] and Von Neumann [400] as well
as passages from more recent works such as those by D. Hofstadter [401] and
de Regt and Dieks [402] concerning the intuitive and visual representability
of physical theories. The questions asked required to perform an author by
author analysis of the content of each of the proposed passages, to compare
the positions expressed by different authors, and also to expose a personal
view of the problem, based on a metacognitive reflection on one’s own way
of learning, understanding and visualizing physics.
The questionnaire was proposed as the beginning of the quantum physics unit.
The first part was compiled in the classroom, while the second part was given
as a home activity. In the next lesson a collective two hours discussion was
carried out, to which students were encouraged to contribute by sharing their
answers and reflections, and developing arguments to sustain their epistemo-
logical positions, in such a way to increase both the individual and collective
epistemological awareness. With this activity, students became aware of the
problem of defining an intuitive content for physical theories; developed a vo-
cabulary to discuss the problem; and perceived that they were allowed to take
their own personal position on the subject.
After its initial introduction, the thread concerning the possibility of defining
an intuitive content for physical theories, and the role of images, analogies,
and mathematical models in doing so, emerged several times during the se-
quence. In particular, it was touched in the second two hours open discussion
activity which was added to the central part of the sequence, which will be
extensively described in the next subsection, as it is more focused on the crite-
rion of multi-perspectiveness. It was also a central theme in the argumentative
paper proposed to students at the end of such discussion. Finally, the subject
resurfaced in the final test-questionnaire (Appendix A.3.4) and in the inter-
views with selected students which were performed at the end of the sequence
(the protocol is reproduced in Appendix C).

5.5.3 Multi-perspectiveness
Multi-perspectiveness is defined in Ref. [34] as the requirement that the physics
content is analyzed from more than one different perspective. For example, in
their work on thermodynamics, the authors approached the subject matter
from both a microscopic and macroscopic point of view.
Multi-perspectiveness and multi-dimensionality are partially overlapping con-
cepts, in the sense that, when the disciplinary content is discussed under two
different points of view, if, as it often is the case in quantum physics, differ-
ent physical perspectives bring about a different epistemological content, the

127
 5. The design of our proposal

introduction of multi-perspectiveness is inseparable form the introduction of

multi-dimensionality. In this sense, in the following we will describe the central
discussion activity we added to our sequence as a means to operatively include
multi-perspectiveness; but it is clearly related to multi-dimensionality also.
Although this criterion had not been systematically taken into account in the
design of our proposal, a few of the physical concepts introduced were indeed
treated under different perspectives. For example, the uncertainty principle
was presented as the expression of a limitation on the amount of information
available in a quantum system, and as an explanatory principle for phenom-
ena such as the tunnel effect, and the existence of a non zero ground state
energy for bound systems. Or, the sum over paths method underwent a pro-
gressive generalization, from the idea of possible paths of a single particle only
constrained by physical limitations, to paths which are possible according to
available information, to indistinguishable histories involving a whole system,
composed of many particles.
To these germinal elements, which were however important for the design of
the final interview protocol, new perspectives were added to the discussion
of the uncertainty principle and wave particle duality, which took the move
from their subsequent historical interpretations (as evolving meaning schemas
[364]) and was intended to present the modern point of view as a synthesis and
point of arrival. Operationally, this consisted in another guided discussion of
one hour time, which took place after the first step of the sequence devoted
to conceptual and epistemological themes (Sec. 5.3.5). The activity revolved
around the following main historical debates and points of view, introduced
and explained by the teacher:

ˆ The Bohr-Heisenberg debate on the meaning of the uncertainty principle

and Heisenberg’s now surpassed explanation of the inequality as due to
a physical perturbation of the system caused by measurement (i.e. the
“Heisenberg microscope”).

ˆ Bohr’s perspective on duality as a more general and fundamental concept

than it appears in the formulation of quantum physics, finding application
also in other disciplines such as linguistics ([403]). This idea was compared
to the position of Schrödinger’s, who was more akin to giving to quantum
objects an entirely wave-like interpretation; to Feynman’s view as expressed
in his own words; to Lévy-Leblonds analogy between the quantum object
and the platypus, which essentially expresses the idea that wave particle
duality is a problem of semantic definition; and finally also to Einstein’s
skeptical view of quantum paradoxes, as embodied in the historical Bohr-
Einstein debate.

ˆ The Schrödinger-Heisenberg debate on the visualizability (anschaulichkeit

[404, 405, 406]) and the intuitive content of quantum physics. The introduc-
tion of this historical theme had of course also the objective of making the

128
5.5. Rethinking the sequence for the 2015 test in high school

general thread described in the previous section reemerge in the discussion

with students.

After this activity, students were assigned the writing an argumentative paper
whose general subject was “Quantum physics and the representation of reality:
possible pathways for imagination and intuition.” In writing this paper, which
could alternatively take the form of a newspaper article or a short essay11 , stu-
dents could take inspiration from short passages by Bohr, Schrödinger, Heisen-
berg, Einstein, Weiss, Lévy-Leblond, and Dirac (See Appendix A.3.2), which,
besides the general subject reported above, often concerned the interpretation
of wave particle duality. Thus, the recurring theme of providing an intuitive
representation of physical theories, discussed in the previous section, and a
multi-dimensional approach to the problem of wave particle duality were in-
tertwined in this activity.
The problems of the interpretation of wave-particle duality and the uncertainty
principle were also a main subject of the final test-questionnaire (Appendix
A.3.4). Finally, the interview protocol (Appendix C) for the final interviews
to selected students also included key questions on comparing different ap-
proaches to the uncertainty principle, the sum over paths concept and wave
particle duality.

11
The guidelines given to students for writing the paper were modeled on the general
structure of the writing and composition test which in Italy constitutes the first task of the
final exam of high school courses. The activity was performed in tight collaboration with
the teacher of Italian language and literature.

129
 5. The design of our proposal

130
Chapter 6
Implementation and results

In this chapter we report on the three experimentations of the teaching-learning

sequence which were performed. One was performed in 2014, with a group of
student teachers (ST) of the University of Pavia; and two in 2015, using the
revised version of the sequence, one again with ST of the Pavia University, and
the second one in a terminal class of secondary school (18-19 years old). For
all three the experimentations we report a) description of the implementation
context; b) sources of data and method of analysis, and c) results.

6.1 2014 test with student teachers

6.1.1 Context and organization of the study
The main test of the sequence in 2014 was carried out with a group of 12 ST
preparing for service as high school level educators. All of them were either
currently working as teachers, or had done so in the past, but they lacked a for-
mal qualification. The group was inhomogeneous: most of the members were
non-physicists (mathematicians or engineers), and their background in physics
varied. Three out of 12 ST had never previously had any formal training in
modern physics; 5 had followed one college course, 4 had taken more than one.
Three of them had previously taught wave optics in high school. We proposed
ST a pre-test, in which we asked mostly open questions. In particular, we
asked to name differences between quantum and classical physics [137]; to in-
terpret the phenomenon of two slit interference in terms of photons; to explain
the meaning of the uncertainty principle, and of wave particle dualism.
The course lasted 8 hours divided in sessions of two hours each, plus a final
exam. Between the third and final session, home exercises were proposed,
which included both formal aspects (i.e. calculation of probabilities of detec-
tion) and more conceptual issues. The exercises were to be solved collectively,
through an online discussion. The result was a very lively and participated
exchange, from which we took many positive indications.

131
 6. Implementation and results

In the last session, we proposed a post test, which contained some of the pre
test questions, some items inspired by existing conceptual repositories on the
understanding of quantum mechanics [121, 122, 123, 140] and quantitative ex-
ercises similar to those previously assigned. The final exam, which included
both multiple choice and open response questions, was performed about one
month after the end of the course.

6.1.2 Sources of data and method of analysis

In this experimentation we collected data from the pre- and post- tests, from
students’ comments on online discussions, solutions to proposed exercises, and
from their final exam. Open response items were analyzed in two different,
although interconnected, direction: to test conceptual understanding, we clas-
sified responses using a phenomenographic approach; while to the aim of de-
tecting signs of appropriation, we mainly analyzed the language, vocabulary
and idiosyncratic expressions used by ST in their answers.

6.1.3 Results
6.1.3.1 Pre-test
The main indications we drew from the initial pre-test can be summarized as
follows:
ˆ As expected ST were initially extremely confused about conceptual issues
regarding quantum physics. In particular, most ST were unable to name
differences between classical and quantum physics, beyond the very vague
idea that the latter concerns the microscopic world. The only significant
diverging answer was given by a student, who wrote that the main difference
lies in the fact that quantum physics doesn’t allow to represent the concepts
it talks about. None of ST was able to provide a satisfying definition of the
uncertainty principle. 11 ST out of 12 could not write anything meaningful
about an interpretation of the two slit experiment in terms of photons; only
one student provided a partial answer writing that the interference pattern
can be interpreted as distribution of probability. Regarding wave-particle
duality, 7 ST seemed to believe that it referred exclusively to a property of
light, and 3 did not answer the question.
ˆ Results from the multiple choice questions on common student wrong con-
ceptions, on the contrary, were not negative. 8 ST correctly answered that
the uncertainty principle holds for macroscopic objects, although its effects
are of negligible entity, and 9 of them correctly ticked “false” on the question
asking whether electrons follow sinusoidal patterns while revolving around
the nucleus.
Summarizing, we may say that the ideas of ST at the beginning of the course
were extremely vague. However, since all of them had a university degree in a

132
6.1. 2014 test with student teachers

scientific discipline, and most of them were teaching mathematics and physics,
they did not hold elementary misconceptions that are reported to appear in
students at their first exposition to the ideas of quantum physics.

6.1.3.2 General understanding of computational aspects

One of the points that were tested through the final questionnaire was whether
students could use the sum over paths approach and phasor method to actu-
ally compute probabilities of detection in simple cases. One of the post-test
items (reported in Fig.6.1) required to compute the ratio of the probabilities
of finding the photon at two detectors W and Z after a “double two slit” setup.
The question was similar to a home assigned exercise, which, however, was

Figure 6.1: Post test item requiring ST to compute detection probabilities.

formulated in open form and had different numerical values. 9 out of 12 ST

provided the correct answer to this item, thus demonstrating that students
acquired a good mastery of the sum over paths method, at least in the sim-
plest cases. The post-test also contained a very similar question, in which the
ratio of probabilities had to be computed by taking into account the acquired
information that the photon had not been detected by a detector present on
slit D. This item was meant to test both technical aspects and the conceptual
understanding of the role of which way measurement in quantum physics. 11
ST provided the correct answer, thus corroborating the conclusion, which will
be discussed later, that students also reached a good level of appropriation of
this conceptual issue.

133
 6. Implementation and results

6.1.3.3 Interpretation of single photon interference

In one of the home assigned exercises we asked students to explore the Mach-
Zehnder simulation (Fig.5.13) and analyze it in terms of single photons, ob-
taining in general quite satisfactory explanations. In the post-test we proposed
a question very similar to the pre-test item concerning single photon interpre-
tation of two slit interference (see Fig.6.2) and this time the outcome was quite
good: most ST provided accurate descriptions and showed a noticeable con-
fidence in talking about photons. To be more precise, 8 out of 12 ST (67%)
discussed the proposed exercise in terms of different possible paths of an indi-
vidual photon; 2 (16%) provided an uncertain description, writing in general
about “phase differences” without attributing them to paths, possibly having
in mind a wave picture; 1 (8%) displayed an incorrect model of two photons
interfering one with another, and 1 (8%) more did not answer the question.
Examples of satisfactory answers include:
“The presence of the material introduces a phase shift in the paths of the pho-
ton passing through A, so that they are out of phase with the paths passing
through B and this changes the probability distribution on the screen”
“In terms of paths, all possible paths passing through A experience a phase
change of π when the film is applied. So, while in the first experiment the
vectors were in phase and the probability of detection was maximum, in the
second the vectors are in phase opposition and give probability P = 0.”
Indeed, during the online discussion already it had become clear that most

Figure 6.2: Post-test open question on single photon interpretation.

ST were acquiring a correct attitude of imagining, and referring to, a single

photon taking all possible paths and interfering with itself.

6.1.3.4 Wave particle duality and the role of measurement

As discussed in Section 5.1.1, one of the most important objectives in the de-
sign of our sequence was to lead students, by means of the Feynman approach

134
6.1. 2014 test with student teachers

to build a consistent and acceptable mental model of wave particle duality and
the role of measurement. Data on this subject was collected through a multi-
ple choice item in the post-test[123], reported in Fig.6.3, and an open question
in the final exam. The results from the multiple choice question were quite

Figure 6.3: Post-test multiple choice question on the role of measurement.

satisfactory: 11 ST provided the correct answer (b), while one answered (c),
signaling a “realist” position.
The open question asked to discuss the role of measurement in quantum physics
in comparison to classical mechanics, and referring eventually to the case of
the ZWM experiment, which had been introduced during the course. Answers
produced by ST were, with almost no exception, extremely satisfactory, dis-
playing conceptual insight and a precise and secure use of language. Practically
all students, including those who had never previously been trained in mod-
ern physics, stated the problem correctly, and showed good understanding of
the central message of the experiment: for indistinguishable alternative histo-
ries, leading to the same result, amplitudes must be added, and interference
is produced. On the contrary, for alternative processes which are distinguish-
able, because which way information has been recorded or is by any means
available, interference does not occur. In one case only the ST answer con-
tained an element signaling a possible underlying deterministic conception:
in this case, while producing an overall correct analysis, the ST sometimes
wrote about “probable trajectories” rather than possible paths. A majority of
students elaborated on the idea that, in the Feynman picture, acquisition of
information on the system restricts its possible paths to only those compatible
with the acquired information (which is equivalent to the wave function col-
lapse idea); and convincingly made the point that the very fact of acquiring
information about the system, and not some sort of disturbance, should be
deemed responsible for the restriction in the possible paths.
In this case also, we report some excerpts from the answers of ST:
“The very moment that information is acquired about the system through a
measurement, part of the possible paths are eliminated. Such information, and

135
 6. Implementation and results

as a consequence such limitation, modifies the experimental outcome.”

“Such information destroys the interference pattern, which is no longer ob-
served. This is because now there is only one possible path arriving to the
detector, and I will have only one phasor to consider.”
“I believe that this is one of the most striking aspects of quantum physics. It
is not necessary that an observer physically perturbs the experimental setup.
It is simply the acquisition of additional information, of “elements of reality”,
which causes a reduction of the possible paths for the photon.”
A complete analysis of the answers of ST shows several linguistic signals of ap-
propriation: students often speak in first person, identifying themselves with
the observer performing the measurement, or with the system itself; they use
sometimes a colloquial language, producing personal terms to express concepts,
different from those introduced during lessons (a “perfectly transparent” ob-
server; acquired information altering the ”universe of possible scenarios”, the
acquisition of ”elements of reality”), and in general they appear satisfied with
the internal models they have built. Thus, from this first test already we can
confirm our initial hypothesis that the language of Feynman paths is appro-
priate for leading student to build a consistent mental model of wave particle
duality and the role of measurement.

6.1.3.5 The uncertainty principle

Both the pre- and post- test contained identical open questions requiring an
explanation of the meaning of Heisenberg’s principle. In Tables 6.1 and 6.2 the
results from the pre- and post- test are reported, grouping answers in classes of
similarity. Reading the tables, one sees the appearance of 5 essentially correct

Table 6.1: Answers to the pre-test concerning the uncertainty principle.

Explanation Number of ST
Limitation on simultaneous measurements of position and ve-
locity (or momentum)
5
Microscopic systems are disturbed by measurement 2
Impossible to know position and velocity of a particle with the
1
same precision
Impossible to determine with precision the position of an elec-
1
tron around the nucleus
Introduction of a probabilistic element 1
No answer 1

analyses in the post-test; also, answers stating that the origin of uncertainty is
a perturbation due to measurement have disappeared. However, some critical
points remain. At least 3 ST give almost identical incorrect or partial answers
in the pre- and post-test, and four retain the somewhat controversial idea [407]
that the uncertainty principle expresses a limitation on joint / simultaneous

136
6.2. 2015 test with student teachers

Table 6.2: Answers to the post-test concerning the uncertainty principle. Es-
sentially correct answers are in bold font

Explanation Number of ST
Limitation on information in principle available on the
5
state of a quantum system
Limitation on simultaneous measurements of position and mo-
2
mentum (or of complementary variables).
Unclear or imprecise (probably referring to a limitation on si-
2
multaneous measurements of complementary variables)
Unclear or imprecise (probably referring to a limitation on in-
1
formation in principle available on a quantum system)
Limitation to position measurement only (of the electron in an
1
atom)
No answer 1

measurements. One ST still does not provide any answer. We conclude that
concerning this conceptual point our results were less than optimal, and in
future implementations our sequence can be improved in this respect.

6.2 2015 test with student teachers

6.2.1 Context and organization of the study
At university level, the main test of 2015 was performed with a group of 19
ST again in the context of a teacher training course. In this case, among
the ST were 2 physicists, of which one had obtained a Ph.D.. Among the
non-physicists, 4 ST had a Ph.D. either, one of which in a subject related to
physics. The majority of the group, however, was composed of mathematicians
who had received very little or no instruction about modern physics. To be
precise, 12 ST had not previously followed any course on modern physics, 5
one course only, and 2 (the two physicists) more than one.
An important difference between the ST groups of 2014 and 2015 was that, due
to the peculiar characteristics of the Italian teacher training system, the 2015
group had sustained a selective initial exam to be admitted to the course (only
about 25% of the participants to the exam were finally admitted to the course),
while the 2014 group had been automatically enrolled. As a consequence, the
2015 group was, on average, composed of people coming from rather brilliant
university careers. Thus, the 2015 group was inhomogeneous with respect to
previous instruction about modern physics, but also was scarcely comparable
to the 2014 group in terms of general level of preparation. For this reason, we
did not carry out any detailed comparison between the results of these stu-
dents and those of the previous years, and did not propose exactly the same
questions in the final test. In this study the focus was on identifying possible

137
 6. Implementation and results

causes of misunderstanding, or interesting questions from students, through an

analysis of lessons recordings. Other objectives were to confirm the extremely
positive results obtained the previous year on the understanding of wave par-
ticle duality and the role of information, and to progress on the medium term
objective of confronting our students with general purpose repositories on the
understanding of quantum physics. As the final exam concerned an entire
course which had included several other areas of physics, not all the subjects
covered in the quantum physics part could be tested.
The available time for the quantum physics sequence was 10 hours. We devoted
the first 2 hours to experimental activities in the laboratory, and developed
the in-classroom part of the sequence in the remaining 8 hours, divided in
two hours sessions. During the first lesson we gave students an initial mini-
questionnaire, which was considerably shortened with respect to the previous
year pre-test, in order to save time for the sequence. The activity on the mea-
surement of the Rydberg constant was assigned as individual home work at the
end of the course, using photos of the spectrum of the hydrogen lamp which
were taken during the first lab session. After the end of the lessons, the final
exam was given, which included, referring to the quantum physics part only,
four multiple choice and one open question.

6.2.2 Sources of data and method of analysis

We collected data from the initial questionnaire, from the sheets of the final
exam, from online discussions and solutions to proposed exercises, and from
the recordings of lessons. The lesson recordings served mainly to provide us
with a reliable feedback on the difficulties which ST encountered during the
sequence, represented mainly by their questions and interventions. The an-
swers to the open response question were analyzed using a phenomenographic
approach and considering linguistic elements as in the previous year test, and
also through the development of a Knowledge Integration rubric [96, 98, 102].
The main reason for using such instrument was that we not only wished to test
students’ understanding of individual concepts, but also felt necessary to study
whether such models could provide them with an integrated, unifying picture
of quantum physics, as opposed to having fragmented models, consistent only
in the context of a phenomenon or experiment. A secondary objective of such
analysis was to have a quantitative measure of the consistency of ST’s answers
in view of a comparison with the results of high school students in a similar
test question, see Sec. 6.3.4.3.

6.2.3 Results
6.2.3.1 Data from the initial questionnaire
In this very short questionnaire, which had to be compiled in 15-20 minutes, we
only asked st (1) whether they had already followed courses in modern physics,

138
6.2. 2015 test with student teachers

and if they had, how many; and (2) to discuss the differences they could think
of between classical and quantum physics. Answers to the first question have
already been accounted for in Sec. 6.2.1. Concerning the second question, as
in the previous year course, students appeared overall quite confused. Three
of them could not name any differences with classical physics, and four more
mixed up quantum mechanics with relativity, believing that the former deals
with objects moving nearly at the speed of light. However, several students
also mentioned relevant differences, such as the uncertainty principle, or the
quantization of energy; often their formulation of such concepts was imprecise
but this was not relevant at this stage. The recurring initial ideas about the
differences between classical and quantum physics are grouped in conceptual
categories and summarized in Table 6.3.

Table 6.3: Differences between classical (CP) and quantum physics (QP) men-
tioned by students in the initial questionnaire. Only concepts highlighted in
more than one answer sheet are shown

Total number of
Concept
occurrences

Quantities which are continuous in CP may be discrete in QP 7

Heisenberg’s principle has no counterpart in CP 6
QP deals with microscopic objects 4
CP is deterministic, QP is probabilistic 4
QP deals with objects moving nearly at the speed of light 4

6.2.3.2 Data from lesson recordings

An analysis of the recording shows that, although the timing was tight, lessons
were very participated. In particular, a substantial amount of time was de-
voted to lively collective discussions during the introduction of the photon and
its properties (about 10 minutes), after the presentation of the Zhou-Wang-
Mandel experiment and its consequences (20 minutes), and in the step devoted
to the Hong-Ou-Mandel experiment and the different behaviour of photons and
electrons (10 minutes). Students posed valuable questions and expressed in-
teresting doubts at other times also. In the following we summarize the main
points that were raised, in a rough chronological order with respect to the
sequence development.

ˆ Interference is due to collisions between photons. In the initial stages of our

sequence, after introducing the photoelectric and Compton effects and the
basic properties of the photon, we asked to students how they could explain
interference from a corpuscular perspective (we pleaded the two physicists
to remain silent). The question was aimed at testing our hypothesis that
the following steps (i.e. showing two slits interference realized with one

139
 6. Implementation and results

photon at a time, and the Grangier experiment) were necessary in order

to avoid the insurgency of hybrid conceptions. In fact, one of the students
immediately elaborated the conjecture that interference is due to photon-
photon interactions, i.e. collisions. Several other ST followed him along this
route; and indeed when we asked what they expected to see if the two slit
experiment was performed throwing only one photon at a time towards the
slits, they all agreed that they expected a very different pattern to appear.
When, immediately afterwards, we showed the video of the experiment, all
these ST agreed that, with this new evidence, the explanation based on
photon collisions no longer seemed plausible.
ˆ Can the vector associated to each path be identified with (a component of)
the electric or magnetic field? This sophisticated question was asked by a
student with a degree in engineering and a Ph.D. on a subject related to
electromagnetic fields. We believe that the best answer to this question is
a definite “no”, for both educational and theoretical reasons. First of all,
the model we are introducing is meant to be generally valid for all quantum
physics, and as such it is evident that the vector has to be interpreted
as the representation of an abstract, complex valued amplitude function.
Furthermore, even in scalar optics, which is the wave analogue of Feynman’s
photon model, the wave function uω (x) satisfying the Helmholtz equation
for a given frequency ω is not simply a component of the electric or magnetic
field; it is instead, in this case also, a complex valued function, bearing a non
trivial relation to the physical fields, whose square modulus is proportional
to light intensity (see Ref. [318] for a full derivation and discussion). The
field components are solutions of the same equation, but are real valued
functions.
ˆ Isn’t all this just a way of representing waves? This question was asked
by a student who was initially rather skeptical, after the explanation of
Feynman’s model of the photon following all paths. To provide an answer,
after recapitulating all the analogies between scalar wave optics and Feyn-
man’s model of the photon, we pointed out the differences, which, at least
at the beginning, mainly concern the language used and the interpretation
of results. We told the student that books sometimes choose to talk about
“probability waves” [408] (which should more properly be called “amplitude
waves”) but this choice brings about the necessity of referring to collapsing
waves in the context of measurement; perhaps she would have appreciated
the usefulness of Feynman’s language later in the course (which she even-
tually did).
ˆ Saving classical probability from failure. As mentioned in Section 5.3.5,
at least two students with a degree in mathematics found the following
exchange illuminating. The dialogue took place after showing, through a
simulation, that the probability distribution for the detection of a photon
in a two slit experiment is incompatible with the (normalized) sum of the

140
6.2. 2015 test with student teachers

probability distributions resulting from diffraction, when either slit is closed.

Teacher: “What would probability theory predict?”
ST: “It would predict that the two slits probability density is the averaged
sum of the two single-slit densities, provided the two events are mutually
exclusive.”
Teacher: “Ok. Then you see a way out for making sense of this experiment?”
ST: “Yes. Assuming that the two events are not mutually exclusive; or in
other words that it is possible for the photon to pass through both slits at
once.”

ˆ In the Zhou-Wang-Mandel experiment interference is between two photons.

This is a pitfall that must carefully be avoided in the presentation of the
ZWM experiment. Some students initially misunderstand the situation,
interpreting the setup as meant to demonstrate interference between two
different photons, one emitted from each crystal. It must be clarified right
from the start that interference in this experiment happens between two
possibilities [409], two indistinguishable alternative histories of the system.
In one history crystal “A” emits the couple of photons (of which only the
“primary” one goes to the detector where interference is revealed), in the
other history crystal “B” does. It never happens (or almost never, and
anyway the rare cases in which it does are irrelevant to the experiment)
that both crystals emit a couple of photons at the same time, and avoiding
such occurrences is precisely the reason why the crystals are chosen as low-
efficiency ones. If well understood, this experiment is extremely illuminating
in demonstrating the effects of measurement without physical interaction,
but time must be taken to thoroughly describe the details of the setup and
to dissipate students’ doubts.

ˆ Information acquired by whom? This question was asked by a rather bril-

liant student with a degree in mathematics, who proposed the following
example: if in the ZWM experiment, when the two alternative possibilities
are rendered distinguishable, if the detector for secondary photons (which
permits distinguishability) is covered with a blanket so that it cannot be seen
by the experimentalist, will interference reappear at the “primary” photon
detector? The answer is definitely “no” in this case, and the basic explana-
tion is that information does not need to be acquired by a sentient observer,
but to be, in some sense, available. However, the question calls into play
the greatest of all mysteries in the interpretation of quantum physics: the
“problem of measurement” in its strictest sense [410, 411] as expressed in
the paradox of Schrödinger’s cat. At what level in the measurement pro-
cess, exactly, does the wave function collapse? Is it at the first interaction
with the measurement apparatus (which, as Bohr thought, must be neces-
sarily modeled as a classical object), or a large enough number of couplings
with the environment must take place? Even so, what is the mechanism
that allows to select only one of the states of the measurement apparatus,

141
 6. Implementation and results

from the correct basis to describe the classical world1 ? These questions
have been debated for nearly a century now, no definitive consensus has
been reached, and certainly we will not try to provide an answer here, nor
even to review an immense literature. But, there is perhaps a consideration
of a more sociological nature that may be useful to students: scientists in
the field of quantum optics, who work on designing experiments and ap-
paratuses at the frontiers of quantum physics, in general are not primarily
trying to understand how and when the possibility of experimentally distin-
guishing two processes destroys interference, but rather they are reasoning
the other way round, taking interference as a measure of indistinguishabil-
ity [412, 413]. For example, if they need to have indistinguishable photons
available, they set up an experiment of the Hong-Ou-Mandel type: if they
get the expected interference effect between alternative processes, they in-
terpret this as evidence of having obtained indistinguishability; but when
the visibility of the interference phenomenon is reduced, or disappears, they
assume that the photons have been made distinguishable, or partly distin-
guishable. In this last case, more often than not, they have no clue as to
how in practice the two processes (and photons) could be distinguished;
they simply assume that some sort of coupling with the environment exists,
mediated by entanglement with other photons, phonons, or electrons, which
has not been properly controlled [414]. In other words, physicists work as
if the “availability of information” on the distinguishability of two processes
is simply defined as the degree (visibility) of their interference effects. A
recent proposal to independently define an objective concept of availability
of information in a way that is consistent with the above use is the program
of “quantum darwinism”[362, 363].

ˆ Why do only bosons and fermions exist? This question was asked after the
introduction of the generalized Hong-Ou-Mandel effect. A possible answer
is of course that this is simply an experimental fact. A more advanced argu-
ment is given by the mathematical consideration which we have reported in
a note in Sec. 5.3.7: the multiplicative factor assigned to the transformation
T which exchanges the roles of two identical particles can only be either 1
or −1, because applying T twice must give the identity. This explanation
can be satisfactory for pre-service teachers with a degree in mathematics,
but whether high school students can accept the argument in some form is
still to be seen, since this part of the sequence was not tested in high school.

ˆ What is the meaning of a “source” and “detector” for the particle in a box?
In the treatment of bound systems using the time independent sum over
paths approach, the source-to-detector terminology is a bit stretched. In
fact, one student asked why the particle isn’t immediately stopped at its
first encounter with the detector, rather than being able to perform an
1
For a review of the successes and shortcomings of the “decoherence” program, which is
alluded to in these lines, see Ref. [278].

142
6.2. 2015 test with student teachers

infinite number of round way trips in the well, and whether this means that
the detector is imperfect. Another ST was confused by the meaning of a
“source” in the case of a particle in a box. We believe this question can
only be clarified in terms of information available on the system: what we
are assuming here, is that the particle was first found at xS , and that an
intrinsically unknown amount of time later is found at xD , with its energy
fixed at the value E. We have no information whatsoever about the time
intercurring between the two events2 ; and such uncertainty must be regarded
as non epistemic, i.e. we must consider paths which need any amount
of time to reach xD starting from xS . We recognize, however, that this
can potentially be a significant terminological problem in teaching quantum
bound systems using this new method. A possible alternative is to speak of
“initial” and “final” positions, but this has shortcomings; in fact, such terms
may appear to allude to a particle starting and ending its history at definite,
albeit not specified, times. In this respect, in particular the term “source”
appears more appropriate as it reminds of a stationary situation, in which
the particle may be created at any time. A possible alternative could be
to speak of a “source” and a “detection event”. More teaching experience is
probably needed to decide about this subtle linguistic issue.

ˆ In the case of the infinite square well, why does the probability of finding the
particle depend on position? Some students misunderstand the constructive
interference condition producing the allowed energy levels for the particle in
a box, interpreting it as the requirement that the vectors corresponding to all
the possible paths are in phase. Thus they are surprised that the probability
of finding the particle depends from its position in the well. However, the
quantization condition Eq. 4.44 corresponds to the requirement that all the
paths which differ for a full orbit in the well are in phase. This leaves the
possibility open that, by changing the positions of the source and detector,
the four basic paths in Fig. 4.4 add up with different reciprocal phases,
producing a variable amplitude vector that, for some choices of the source
and detector positions, may even be zero. The important point to stress here
is that the allowed energy levels are those values for which the probability
of finding the particle is not identically zero everywhere.

6.2.3.3 General understanding of computational aspects

ST of the 2015 course were able to solve problems at least as well as their
counterparts of the previous year, and probably better. We assigned to them
a significantly higher number of home exercises (see Appendix B.2) and they
required less online assistance from their tutor in solving them, producing in
general correct and well documented solutions. Given these encouraging data,
we had less necessity of testing their general comprehension of the basics of the
sum over paths algorithm; in the final exam we only required them to solve one
2
Except that we know their time ordering, i.e. that the first event precedes the second.

143
 6. Implementation and results

exercise of this kind, presented as a multiple choice question, which is reported

in Fig. 6.4

In the Mach-Zehnder interferometer represented

in figure, the source emits a single, monochro-
matic photon. The two arms of the interferometer
(BS1 − M 1 − BS2 and BS1 − M 2 − BS2) have the
same length, as do the traits BS2−Detector A and
BS2−Detector B. Consider detectors with 100%
efficiency. In these conditions, it can be assumed
that a photon will be detected:

a. Never, neither at detector A or B.

b. Certainly or almost certainly at A.

c. Certainly or almost certainly at B.

d. At A or B with 50% probability.

e. At both detectors a photon will be detected

with half the original energy.

Figure 6.4: Multiple choice question of the final test concerning the Mach-
Zehnder interferometer.

The question was answered correctly by 16 ST out of 19. Several students

included in the sheet an unrequested, detailed analysis of the possible paths
considered, phase differences accumulated from reflections, and the calculation
of amplitudes and probabilities, displaying confidence in treating the problem.
Of the three ST who gave a wrong answer, 2 chose the alternative (d), while
one chose (a). In one of these three cases, based on the accompanying analysis
reported on the test sheet, the error was not due to a global misunderstanding
of the sum over paths algorithm, but to a mistaken count of phase losses due
to reflection. In the remaining two cases the wrong answer contained no addi-
tional comments. Results from this item confirm the impression we had during
the sequence development, that students had reached a good understanding of
the basics of Feynman’s method.

6.2.3.4 Energy of emitted photons in atomic transitions

One of the multiple choice items of the final exam concerned the understanding
of energies of emitted photons in atomic transitions, and was modeled upon
similar ones contained in the QMCS 2.0 repository [122]. Such questions are
meant to test the known possible wrong conception [176, 183] that the energy
of an emitted photon corresponds to the energy of a level, rather than to the
difference between the energies of two levels. The question is reported in Fig.
6.5. In this case, 18 ST out of 19 answered correctly the question, with only one
ST picking alternative (e). Thus, the fraction of correct answers was 95%, a
result which may be compared to the 64% of correct answers for the analogous

144
6.2. 2015 test with student teachers

An electron in an atom can occupy the energy levels

shown in the figure. The electron is initially in the
state with energy E3 . What is the highest energy
that an emitted photon can have in this situation?

a. E4

b. E4 − E1

c. E3

d. E3 − E1

e. E3 − E2

Figure 6.5: Multiple choice question concerning the energy of the emitted
photon in an atomic transition

item (3) of QMCS 2.0 reported in Ref. [122] in a test with N = 370 students of
modern physics, majoring in either physics or engineering. Although it must
be reminded again that our sample was composed of graduate students who
had passed a selective exam, yet most ST were at their first experience with
modern physics. We may conclude that our students correctly understood the
basic rules of photon emission and absorption, and did not acquire the mistaken
conception of identifying the energy of the photon with the one of a level. This
successful result may be connected to the theoretical and experimental work
on the Balmer series, the observation of discrete spectra, and the measurement
of Rydberg’s constant.

6.2.3.5 Wave particle duality and the role of measurement

A very important objective for the 2015 test was to confirm the positive re-
sults obtained the previous year on student’s understanding of wave particle
duality and the role of measurement. In the final exam we tested these aspects
through two multiple choice and one open question.
The first multiple choice item was the same assigned in the 2014 pre-test (Fig.
6.3). In this case, all students with no exception provided the correct answer
(b). It may be significant to add that several students included further, unre-
quested comments in margin of this question. In particular, 6 ST commented
that the acquisition of which way information reduces the possible paths, or
histories, for the electron, negating the possibility of interference. In a slightly
different perspective, 2 ST noted that the acquisition of information makes the
histories distinguishable, implying a return to the classical concept of proba-
bility. These notes give additional weight to the statement that by using the
Feynman approach students can build satisfying mental models of wave parti-
cle duality.
The second multiple choice question on this subject concerned an hypothetical

145
 6. Implementation and results

case of Mach-Zehnder setup with an interaction-free which way measurement.

The question is reported in Fig. 6.6.
Again, all students provided the correct answer (e) to this item, and many

In the Mach-Zehnder setup of the previous exercise,

a third detector C is added, as in the figure. Assume
that the detection in C is interaction-free, so that
the photon, even if revealed at C, can continue on
its path. Consider detectors with 100% efficiency.
In these conditions, it can be assumed that a photon
will be detected:

a. Never, neither at detector A or B.

b. Certainly or almost certainly at B.

c. At both detectors a photon will be detected

with half the original energy.

d. Certainly or almost certainly at A.

e. At A or B with 50% probability.

Figure 6.6: Multiple choice question concerning the Mach-Zehnder interferom-

eter with interaction-free which way measurement.

of them included additional notes and comments. In 7 cases, these comments

concerned the reduction of possible paths due to the presence of the detector,
with one ST connecting to the idea of wave function collapse. Other 7 notes
were more focused on the idea of distinguishability bringing back the classical
probability rules, with 2 ST adding the consideration that this happens be-
cause the presence of the detector ”renders the two events of the photon taking
the two arms of the interferometer, mutually incompatible”.
The open question was formulated as follows:
”Acquiring information on a quantum system through a measurement often
modifies its future evolution. Comment the previous statement, confronting
it with the classical case and referring to the two slit experiment, or to other
experimental situations which were discussed during the course.”
In this case, one ST left the answer blank. Two students produced partial
or inadequate answers. In one of these two cases, the ST only superficially
described the two slit experiment and the effect of a which way measurement,
without providing explanations, beyond a vague remark on the non neutrality
of measurements in quantum physics. In the second one, the student tried to
explain the same phenomenon by appealing to the uncertainty principle in a
rather confused way. These may be defined as clear cases of non appropria-
tion [34]. The student with a Ph.D. in physics produced a detailed analysis of
the measurement process in terms of a projection of the state of the quantum
object into an eigenstate of the operator associated to the measured observ-
able, which is of course a correct answer, but irrelevant to the present study.
The remaining 15 students satisfyingly answered the question in terms of the

146
6.2. 2015 test with student teachers

reduction of possible paths due to information which is rendered available,

or the related concept that interference only can happen for indistinguishable
processes. The results are summarized in Table 6.4, where the main ideas in-
troduced by ST in their answers are reported, grouped in classes of similarity,
specifying the total number of occurrences of the concept, as well as the num-
ber of answers in which it seemed to assume the primary explanatory role.

Table 6.4: Concepts introduced by ST in their explanations of the role of

measurement in quantum physics

Total number Occurrences

Concept of as primary
occurrences explanation

Available information reduces the possible paths 12 9

Indistinguishable processes interfere, while for distin-
10 6
guishable ones the classical probability rules are used.
Acquisition of information does not necessarily involve
9 0
physical interaction
Non neutrality of quantum measurement 3 1
Explicit generalization to possible histories 2 0
Wave function collapse 2 0
Projection of the state into an eigenstate of the mea-
sured observable
1 1
Uncertainty principle 1 1

As shown in Table 6.4, a significant number of ST also highlighted the idea

that the acquisition of information on a quantum object does not necessarily
involve a physical interaction with the object itself, as demonstrated by the
ZWM experiment. Several students, in particular those explaining the role
of measurement in terms of distinguishability, referred in general to processes
rather than paths; but in two cases only the answer contained an explicit
generalization to histories of a quantum system involving more than one parti-
cle. Concerning the experiments which ST considered in their answers, which
are not reported in the table, all of them treated the two slits case; one dis-
cussed the case of three slits; 6 explained in detail the Zhou-Wang-Mandel
experiment, 4 analyzed the Mach-Zehnder interferometer and one the Hong-
Ou-Mandel setup.
We also analyzed the results using the method of rubrics for knowledge inte-
gration [101, 107]. In Table 6.5 the coding levels and descriptions are shown,
while in Table 6.6 the results are summarized. In this case, the answer by the
ST introducing projections on Hilbert spaces was removed from the sample.
The average score of ST for this question is 3.3, which is very high for this
kind of coding.
A complete reading of the answers of ST under a linguistic perspective con-
firms the conclusions drawn the previous year concerning the presence of clear

147
 6. Implementation and results

Table 6.5: Rubric for scoring accounts of the role of measurement in quantum
physics

Score Level Description

Does not answer the question, or the answer is com-

0 Irrelevant
pletely unrelated to the question.
The concepts introduced in order to answer are wrong
1 Non-normative
or connected in a scientifically invalid way.
Makes a connection between the acquisition of which
way information and the disappearance of the inter-
2 Partial
ference pattern, but does not try to explain or the
explanation is non-normative.
Explains that which way information reduces the pos-
3 Full sible paths, or that distinguishable processes follow
classical probability rules.
As “full”, but connects the explanation to the analy-
sis of at least one physical system in a complete and
4 Complex
scientifically valid way, or provides a suitable gener-
alization to arbitrary indistinguishable processes.

Table 6.6: Number of ST answers for each level of the rubric coding, and
average grade.

Average
Level 0 Level 1 Level 2 Level 3 Level 4
level
1 1 1 3 12 3.3

signs of appropriation. Also, those students who discuss the ZWM experiment
appear even more confident in doing so than the previous year ones, analyzing
experimental details and not only the general conclusions, possibly due to the
larger time for discussion which was left in the course during and after the
introduction of the setup. We report a significant extract from the description
produced by one ST (underlined words are in the original). The student has
a degree in mathematics, and already during the course had declared that the
ZWM experiment had been for him very enlightening:
”The second photon of each couple, instead, follows a different path, and can
be used as a signal to understand whether the first photon has been emitted
from one source, or the other. Surprisingly, the presence of a detector which
does, or does not, reveal the passage of the second photon, modifies the evolu-
tion of the experiment also for what concerns the first photon, which has not
even come near to such detector. Thus it is the acquisition of information, and
not the physical interaction with the observer, to influence the experimental
outcome.”

148
6.3. 2015 test with high school students

6.3 2015 test with high school students

6.3.1 Context and organization of the study
The study was conducted in the class V B of the Liceo Scientifico Enrico
Fermi, in Genova. In Italy, liceo is one of three main branches of the high
school system, and its primary mission is to prepare students to a future ac-
cess to University. Liceo scientifico in particular is oriented towards the study
of sciences, and is the only kind of secondary school in which a significant part
of the final year physics curriculum is devoted to modern physics. Liceo E.
Fermi is not, however, an elite high school: it is positioned in a working class
district in the western part of Genova; the average background of students is
lower middle class; and its student population includes a significant percentage
of first and second generation immigrants, something which is not found in the
licei of the center and eastern parts of town3 , which truly are traditionally
attended by students from the economic and cultural elites. As a consequence,
on average the scientific and cultural capital [415, 416] available to students
from their family backgrounds is not as high as it could be expected for a
school in this category.
Class V B was composed of 18 students, 10 male and 8 female. The group had
a past history of poor results in mathematics and physics, due in part to a
frequent change of teachers during their high school studies. At the end of the
fourth year, all students but four had been assigned a below average grade in
mathematics, which required them to sustain an additional reparation exam
in order to progress to the next year classes. Three students were repeating
the final year for the second time; however they had not previously received
any instruction on concepts of quantum physics since, as mentioned in Section
1.1, the physics curriculum reform for the final classes of high school came into
effect in 2014-2015.
For 2014-15, the courses of mathematics and physics, which in the final class
of Liceo Scientifico comprise 4 and 3 weekly hours respectively, were assigned
by the school Principal to the same teacher4 , in an attempt at providing stu-
dents with a more cohesive view of these two subjects, and improving their
results. This allowed more flexibility in designing the educational intervention
on quantum physics as, if necessary, a few hours could be borrowed from the
mathematics course and returned in a different occasion. The total amount of
time devoted to quantum physics was about 33 hours, divided as follows: 3
hours for experimental activities; 9 hours for tests, questionnaires and the pro-
duction of argumentative papers; 2 hours reserved for open discussion; and the
remaining 19 hours for traditional classroom lessons, including time employed
for the sequence development, for the introduction of historical and epistemo-
3
In year 2013, the percentage of students of foreign nationality in the high schools in
Genova was 10.6% [417]. However, such students are overwhelmingly concentrated in schools
of the western and northwestern areas of the city [418]
4
The teacher is the author of this dissertation.

149
 6. Implementation and results

logical themes, for providing the solution to exercises assigned as homework,

for answers to students’ questions and for discussions spontaneously opened
by students. The time for interviews to selected students was not included, as
we separately performed the interviews with one student at a time, while the
rest of the class was involved in other activities. The timeline for the sequence
development is schematically reported in table 6.7.

Table 6.7: Schematic timeline for the development of the intervention on quan-
tum physics in high school, reporting the duration of each activity.

The photoelectric effect: critical reflections on

16/2 ˆ 1h
Maxwell’s model of light
18/2 ˆ Experimental activities at the university of Pavia 3 h
6/3 ˆ Einstein’s explanation of the photoelectric effect 1 h
13/3 ˆ Compilation of the initial questionnaire 1 h
18/3 ˆ Open discussion on the initial questionnaire 1 h
27/3 ˆ Main sequence development (I) 2 h
1/4 ˆ Main sequence development (II) 3 h
8/4 ˆ Main sequence development (III) 3 h
15/4 ˆ Main sequence development (IV) 3 h
24/4 ˆ Main sequence development (V) 2 h
Interlude on historical-epistemological themes
27/4 ˆ 1h
and discussion
29/4 ˆ Writing of argumentative papers 4h
6/5 ˆ Main sequence development (VI) 3h
11/5 ˆ Final graded test 2h
14/5 ˆ Discussion of some loose ends 1h
3/6 ˆ Final conceptual test 2h
8/6 ˆ Interviews to focal students ∼ 40 m each

When in the timeline in Table 6.7 steps are marked as “Main sequence
development” and a progressive number, by no means it should be intended
that lessons where an attempt to exactly reproduce the development of the
sequence as performed with pre-service teachers. First of all, in many cases
the time devoted to answering students’ doubts on home assigned exercises,
problems, and on the content of the previous lesson occupied a large part of
the available hours. Secondarily, in the spirit of the Pavia approach to the
design of teaching learning sequences (Section 2.1.1) while the core content,
with its principal conceptual correlations and methodological choices, was cer-
tainly preserved, peripheral or cloud material was often adapted, expanded or
reduced taking into account students’ doubts, needs and questions.
For the experimental activities, students were taken on a daily trip to the edu-
cational laboratories of the University of Pavia. All other lessons, discussions,

150
6.3. 2015 test with high school students

and tests were performed in the classroom, which was endowed with a PC and
interactive white board, as well as a traditional blackboard.
All classroom sessions, excluding tests, were audio recorded. Besides the
physics teacher, other persons were involved in the practical realization of
this experimentation:

ˆ A graduating Master student of the University of Pavia was present in class,

in charge of taking notes on the development of activities, and sometimes
with a tutoring role in the solution of exercises. Also he was one of the two
conductors of the final interviews to focal students.

ˆ A researcher of our group guided, together with the teacher, the class in the
experimental activities in Pavia; also, he co-conducted the final interviews.

ˆ The teacher of Italian Language and Literature of Liceo Fermi was involved
in the activity concerning students’ composition of essays; she discussed and
approved the guidelines, and co-supervised the activity.

Since one of the students had a mild learning disability, a supporting teacher
of Liceo Fermi specialized in disabilities was also present in the classroom in
many occasions; her presence is not related specifically to the quantum physics
sequence, as she was assigned to the class for the entire year.

6.3.2 Sources of data and method of analysis

At the end of the sequence, we had collected a considerable amount of ma-
terial to extract data from, including: the recordings of all lessons and dis-
cussions; the initial questionnaires; written solutions produced by students to
many exercises; the graded test, primarily (but not exclusively) composed of
schoolbook-like problems; the final conceptual test (which was not graded);
students’ argumentative papers; and the final interviews. The analysis of such
data is rather complex, and will be organized as follows. In Section 6.3.3 and
the subsections therein contained, we will provide a bird’s eye view of the
progressive buildup of the learning environment and the development of the
sequence, paying special attention to the general response of students, and fo-
cusing more on those activities which are less typical the high school physics
course: epistemological discussions, the production of argumentative papers,
and final interviews. In this section we will not discuss in the detail the con-
tributions of each students, which will be analyzed later. In Section 6.3.4 and
its subsections, an analysis of student’s understanding and learning difficul-
ties will be carried out, both on a qualitative and quantitative level. To this
aim, we will consider data from lesson recordings, exercises, and tests. For the
final conceptual test using open response questions (explanation items), a cru-
cial piece of data analysis will be provided by the evaluation of answers using
Knowledge Integration rubrics, in a similar way as in the previously discussed
studies with student teachers. Finally, in Sec. 6.3.5 and related subsections,

151
 6. Implementation and results

all data sources will be re-evaluated and triangulated, starting from the fi-
nal interviews, in search of evidence of appropriation. We will draw profiles
for selected students based on idiosyncratic, signature ideas expressed during
the interview; and such ideas will be traced back to individual contributions
in written productions and oral discussions, looking for evidence of a global
consistency in the students’ discourse.

6.3.3 General development of the activities and stu-

dents’ response
6.3.3.1 Experimental activities
Students performed activities on the measurement of the Planck constant h
using two different methods (Sec.5.3.2) and on the observation of spectral lines
from gas discharge lamps using a low cost spectrophotometer (Appendix D)
producing photos which were shown again at the end of the sequence, when we
introduced the study of atomic models. They responded well to the laboratory
session, and several of them were even enthusiast of it. In fact, they complained
that in their high school physics course they only had performed experiments in
the first year. However, due in part to the fact that the activities on quantum
physics were performed in the afternoon (in the morning they had conducted
experiments on electromagnetism) not all of them could remain concentrated
till the end. Many of the students retained a very strong impression from
the experiment on the photoelectric effect, which helped them to accept and
motivate the photon concept. Also, the observation of discrete spectra were
very useful for them at the end of the learning path, when the atomic model
was introduced and its consequences discussed.

6.3.3.2 Initial questionnaire and related discussion

As discussed in Section 5.5.2, the objective of the initial questionnaire (Ap-
pendix A.3.1) which was divided in two parts, was to stimulate students’ reflec-
tion on the different possible ways to intuitively understand a physical theory
and, at a meta-cognitive level, how they individually represented and compre-
hended such theories. In particular, the attention was drawn on three possible
instruments, which were the main common theme between the two parts of
the questionnaire: images, analogies, and mathematical models. It must be
noted concerning the representation through images, that the three physical
subjects for the first part of the questionnaire (the Joule effect, the magnetic
force on a current carrying wire, and the second principle of thermodynamics)
were chosen in such a way that likely students would have drawn, for the first
two of them, a direct representation of the physical system; but for the last
one, a picture of a more abstract model, such as a representation of the Kelvin
or Clausius formulation of the principle using schemes of thermal machines
and heat repositories. Indeed, several students produced such a diagram. This

152
6.3. 2015 test with high school students

was not done by chance, but was meant to initiate a reflection on the different
meanings that may be attributed to the possibility of visualizing a physical
theory.
The discussion of the questionnaire was very well received by students, in
part because an entire hour of open discussion is a quite uncommon activity
in the Italian high school system. The ideas of students were collected on
the three possible strategies for understanding physical theories at an intu-
itive level which had been highlighted (images, analogies, and mathematical
models), each connected to a different citation, by Hertz, Maxwell, and Von
Neumann respectively. In particular, Maxwell’s concept of a physical analogy
as something having only a partial correspondence with the system studied,
and whose limits must always be kept in mind, was discussed at length: stu-
dents proposed the examples of the analogy between water flowing in tubes an
electric circuit, and between electromagnetic and water waves, which, they ar-
gued, may be misleading because it suggests that the former need a medium to
propagate. The teacher established a connection with the forthcoming study
of quantum physics, by pointing out that a theory would be introduced, which
is based on a sound mathematical model, but for which it is difficult, and
some would even say impossible, to provide an intuitive representation. In this
context, the decline, which had already started before the birth of quantum
physics, of the validity of Lord Kelvin’s statement “I am never content until I
have constructed a mechanical model of the subject I am studying. If I succeed
in making one, I understand; otherwise I do not.” [419] was also discussed.
When confronted with the question of explaining their personal approach to
understanding physics, as expected most students highlighted the importance
of images and/or analogies, and very few assigned a significant role to the
mathematical model in intuitive understanding. Of course the primary objec-
tive of this discussion was not dividing students in categories according to the
above lines, but to develop an appropriate language to discuss the problem,
and to initiate the social construction of individual epistemological positions.
In this respect, the last part of the discussion was very interesting, because stu-
dents shifted the focus of debate on the role of mathematics and the question
of why is it so successful in interpreting and predicting physical phenomena.
Here, two very different positions were opposed to each other, reproposing a
classical philosophical debate about the nature of mathematics. On one hand,
some students took an intuitionist position, asserting that mathematics is the
result of a human constructive activity, which has the objective of building an
appropriate language to describe facts of the world, and as such, its success in
describing them is a pure tautology. Other students took a platonist position,
defending the idea that mathematics pre-exists human beings, and its suc-
cess in describing nature has to do with an intrinsic order or symmetry in the
laws of physics. The discussion was participated; however, only about half the
students intervened spontaneously; others, only spoke if solicited. This was a
constant feature over all the activities, with the students animating discussions

153
 6. Implementation and results

remaining consistently the same.

6.3.3.3 Main sequence

According to our subjective impressions, students’ reactions to the develop-
ment of the main core of the sequence were not very different from the exper-
imentation with student teachers, except that each step required about twice
the time. Students needed time to dissipate their doubts, to understand how
to solve exercises, to reflect on conceptual issues. Since the way the material
was introduced was very different from the treatment in their textbook, they
were given the slides of the lessons as learning material for home study. They
did not complain about this practice, as it was commonly used also by other
teachers, when they were not satisfied about a particular chapter of section of
a textbook. Students’ questions and misunderstandings that are relevant to
future implementations of the sequence will be discussed in Sec. 6.3.4.1.
Social interactions during discussions in the class were complex and although,
as mentioned in the previous Section, the number of students actively parte-
cipating in discussions remained more or less the same, the roles played by
individual students evolved. Some of these developments are discussed in Sec-
tion 6.3.5.

6.3.3.4 Historical-epistemological themes and central discussion

As discussed in Section 5.5.3, in a central interlude after the discussion of
conceptual themes of quantum physics from the point of view of Feynman’s
approach, different historical-epistemological perspectives were introduced on
the concepts of wave particle duality and the uncertainty principle. Due to
constraints in the school scheduling of time, this was done in two separate ses-
sions, of about half an hour each. Concerning wave particle duality, the teacher
presented Bohr’s, Schrödinger’s, and Einstein’s historical points of view. At
the same time Lévy-Lebond’s analogies of the quanton with the platypus [150],
and with the cylinder as circle-rectangle, were connected with Feynman’s per-
spective. Regarding the uncertainty principle, the Bohr-Heisenberg debate
was discussed. Also, the Schrödinger-Heisenberg controversy on the possibil-
ity of identifying an intuitive, or visualizable, content in quantum physics was
commented. Students were interested by Lévy-Lebond’s analogies, and the
connection between the idea of “quantum object” and Feynman’s model, but
many of them were also fascinated by Einstein’s ideas on the incompleteness of
quantum physics. Thus, a discussion was raised, as some students wanted to
reconsider the two slit experiment to see if different explanations were possible.
For example one student produced the theory of a particle physically contain-
ing other particles, splitting and then rejoining at the detector. Concerning
the uncertainty principle, most of the students who participated agreed that
Heisenberg’s interpretation did not really fit with the modern experimental
evidence which had been presented. Although the activity probably met the

154
6.3. 2015 test with high school students

objective of challenging an authoritative vision of science, the discussion was

a bit confused, and appeared less relevant and focused than the one following
the initial questionnaire. In future implementations, perhaps the positioning
of these debates within the main sequence has to be reconsidered, because
students should have time to reflect on the concept of wave particle duality
from the point of view of Feynman paths, before alternative perspectives on
the same concept are presented. Also, the activity should be given more time,
to permit a more in depth discussion of the objections and positions expressed
by students in relation to the historical debates; or it should be preceded by
a home activity in which students read part of the content on their own, in
order to save time for actual discussion.

6.3.3.5 Argumentative papers

The activity on the production of brief essays or newspaper articles was car-
ried out in a 4 hours time; the guidelines given to students are reproduced
in Appendix A.3.2. The compositions were primarily graded for the course
of Italian Language and Literature, whose teacher preliminarily approved the
guidelines and co-supervised this activity. Students were told that their essays
could have received a grade in physics also, if they had included significant
physical content, and only if the understanding displayed was satisfactory. No
negative grades in physics would be assigned for this activity. This was made to
incentive students to produce arguments grounded in the disciplinary content
of physics, of course remaining within the constraints of the writing style for a
newspaper article or divulgation essay, i.e. no formulas or graphs representing
phenomena or laws.
The activity was well received by students, who were all present, and generally
interpreted it as an opportunity to do interdisciplinary work, and to discuss
physics from a more conceptual, qualitative and philosophical point of view,
also in the perspective of their forthcoming final exam.
The works were generally on par with student’s usual written productions in
the Italian Literature course; the average grade assigned by the Italian teacher
was 6.5/10. Concerning the disciplinary content, we did not see gross misun-
derstandings of the central ideas of quantum physics show up in the compo-
sition, and students did not appear to try to make the theory say something
it doesn’t, or improperly extend its results to unrelated fields, as something
happens in popular accounts. All students but one discussed wave particle du-
ality, which was a central theme in the quotations reported in the guidelines.
Several of them introduced and discussed further physical content, such as ex-
periments or principles, going beyond the guidelines suggestions. Also, some
students made a connection to the ideas of philosophers they had studied, the
most cited ones being Wittgenstein (mentioned in 4 papers) and Popper (in
3). The compositions will be analyzed individually in Section 6.3.5.1, with
the aim of extracting data for the identification of focal students and cases of
appropriation. Here we only display, in Table 6.8, a summary of the global

155
 6. Implementation and results

content of the essays, reporting how many of them discussed the suggested
texts, and other elements present.

Table 6.8: Content of students’ compositions

Quotation suggested in the guidelines Number of citing compositions

Bohr 14
Lévy-Leblond 11
Einstein 11
Schrödinger 10
Heisenberg 9
Dirac 5
Weiss 1
Principle or experiment, independent from the pro-
Number of citing compositions
posed quotations, introduced and discussed
Young’s experiment 7
Uncertainty principle 5
Photoelectric effect 4
Sum over paths formulation 3
Probabilistic interpretation 3
De Broglie wavelength 3
Compton effect 3
Mach-Zehnder experiment 1
Zhou-Wang-Mandel experiment 1
Entanglement 1

6.3.4 Results on student’s understanding and learning

difficulties
6.3.4.1 Data from lesson recordings
As in the case of the experimentation with ST, we consider data from high
students’ interventions during lessons, enumerating some of the most relevant
questions and difficulties which have emerged.

ˆ The photon may split in two / the photon is a mini-wave After the introduc-
tion of Young’s experiment with one photon at a time, students immediately
realized the problem of explaining interference in a corpuscular picture. As
we had anticipated, one student hypothesized that the photon could be
divided in two at the slits. Another suggested that the photons are “mini-
waves”, probably having in mind some kind of “energy lump” conception.
The formation of these hypothesis is precisely the reason why, immediately
after, we showed to student a beam splitter and discussed the Grangier

156
6.3. 2015 test with high school students

experiment, also through its demonstration in an online “realistic virtual”

laboratory [312].

ˆ Why do only “external” reflections lead to phase shift? The rule according
to which some, but not all, reflections of the photon path involve a phase
shift was mysterious for some students, as they did not understand its origin.
The rule is necessary for understanding the functioning of beam splitters in
optical experiments, and its explanation can be provided as experimental
evidence of the interference fringes produced in the Lloyd mirror, or citing
other classical examples in wave optics, such as reflection from multi-layer
thin films. The final justification for the rule comes of course from a micro-
scopic theory of absorption and re-emission of photons (i.e. QED) which is
beyond the scope of our introductory sequence.

ˆ What is the reason for uncertainty? / Uncertainty occurs in formulas, but

is it there in reality? This kind of questions, asked by some students during
the introduction of the uncertainty principle, represent entirely understand-
able resistances in abandoning the classical, deterministic world view. The
ultimate justification of uncertainty can only be given as an established ex-
perimental fact, and its interpretation should be discussed in an historical
perspective, as a complete disentanglement of all its intertwined meanings is
still an active research area. However, we also tried to prevent students from
forming conceptions that are clearly inadequate, e.g. the idea that uncer-
tainty represents an unavoidable experimental deviation from a true value;
and to provide students with the most uncontroversial current definition of
uncertainty, i.e. a limitation on the possibility to prepare an arbitrary state,
or equivalently, a limitation on the information in principle available on a
quantum system.

ˆ Is a quantum object never at rest? One of the students, during the discus-
sion of the uncertainty principle, intuitively grasped one of its most impor-
tant consequences, stating that “this means that a quantum object is never
at rest”. In our answer we noted that one had to be careful in defining the
meaning of “being still”, but we mostly confirmed the student’s intuition.
Later, the question was recalled and rediscussed, much to the student’s
pride, at the point in the sequence in which we computed, by applying the
uncertainty principle, the ground state energy for the particle in a box.

ˆ How can time be uncertain? This question occurred during the discussion
of the time-energy uncertain relation. From a theoretical point of view the
question is appropriate, since time is a parameter in quantum as in classical
physics, and the time-energy uncertainty relation does not share the same
position within the theory as the “proper” relations between non-commuting
operators. It can be proven [420] that an uncertainty relation for energy and
time can be defined if “uncertainty” in time has to be intended as an interval

157
 6. Implementation and results

of time in which observable quantity of the system chance sensibly. How-

ever, the question can also be used as a handle, to explain to students that
the whole treatment of quantum phenomena at fixed energy (for example,
the discussion of the two slit experiment with “monochromatic” photons)
is necessarily made in the assumption of approximately infinite uncertainty
in emission time. In the case of atomic decays, for example, uncertainty
in time has to be understood as mean lifetime of the state; and as a con-
sequence, as it is intended in many exercises, uncertainty on the time of
emission of the photon.
ˆ Where is the paradox in the “measurement problem”? This somewhat un-
expected question was asked during the explanation of the two slit experi-
ment with acquisition of “which-way” information: the student argued that
since the sum has to be done on all possible paths, it was no surprise that,
eliminating some of the possibilities with measurement, the outcome would
change. To a deeper investigation, this remark was actually due to a mis-
understanding of the difference between classical and quantum probability
rules: the student meant to say that it is no surprise that the resulting
distribution of detection events when the photon passes through one known
slit only (say P (A)) is different from the one corresponding to the case in
which it the slit is not known (P (A) + P (B)). In the end, when the student
understood that the distribution of detection events when the two slits are
open and which way information is not known is actually not P (A) + P (B),
he admitted that the result was indeed paradoxical. As we wrote elsewhere,
this point is particularly delicate, and had to be focused through specifically
designed exercises.
ˆ Confusion of formulas valid for photons and electrons. As reported in lit-
erature (see Section 3.1.1.1) some students, both in home and classroom
exercises, incorrectly applied formulas valid for the photon (for example
E = hν) to the case of the electron. We observed this appear three or four
times in total, counting all the exercises solved by students. It is important
to make clear to students that the identity in the behaviour of photons and
electrons observed in the case of the two slit experiment doe not mean that
exactly the same formulas are valid for the two types of quantum optics;
and in particular to highlight the role of mass in determining, together with
velocity, the De Broglie wavelength of a massive quantum object.
ˆ Do the electrons move in wavelike trajectories because they interact with
photons? Investigating the meaning of this question, which a student posed
after the discussion of Young’s experiment with electrons, revealed a com-
plex alternative model, with two intertwined issues. First, the student at
this point believed that electrons move in sinusoidal-shaped trajectories,
which is a well known incorrect depiction (see Section 3.1.1.1). Secondarily,
he believed that such behaviour, for massive quantum objects, is due solely
to the interaction with photons, rather than to their intrinsic properties;

158
6.3. 2015 test with high school students

and so he was convinced that if somehow the two-slit experiment with one
electron at a time had been repeated avoiding interactions with photons,
the interference pattern would not appear. The student could be convinced
with relative ease that “moving in wavelike trajectory” is not at all sufficient
for producing an interference pattern; however, the idea that he should not
attribute a privileged role to the photon only, and that matter has quantum
properties as well, left him more in doubt. In this case, the student clearly
showed resistances in abandoning the depiction of the electron as a small
classical marble.

6.3.4.2 Assigned exercises and school graded test

Students of V B were used to being required to solve as home work practically
all the exercises contained in their textbook and at times some additional
ones. No exception to this rule was made for the quantum physics sequence,
as they were gradually assigned all the exercises of the relevant chapters of
Cutnell-Johnson [193], plus most of those specifically concerning the sum over
paths approach (Appendices B.1 and B.2). Concerning textbook exercises,
students did not seem to find significantly more difficulties than in those of
other chapters; it must however be remembered that the typical high school
exercises on this subject only require understanding the question and correctly
applying a single formula. Whenever doubts arose on these problems, they
were discussed and solved in class. In the additional problems, students were
less successful. Only about half of the class even tried to do the assigned
home work, and only three or four individuals usually succeeded in producing
a correct solution. Some of the students who had tried to solve the exercises
were called to discuss them with the teacher and the rest of the class.
About a week after the end of the sequence, students performed the test meant
to assign a school grade. Such test, which is reported in Appendix A.3.3, was
composed of five exercises, to be solved in 1 hour and 40 minutes time. The
first three exercises concerned the Compton effect, the uncertainty principle,
and the De Broglie wavelength, and were similar to typical textbook items;
however the first two of them were in two parts, and the second part involved
additional mathematical or conceptual difficulties. The fourth exercise, in three
parts, was about the sum over paths approach, and concerned a system which
was new for students: the Michelson interferometer. Part (a) simply required
to identify the possible paths; part (b) required to compute the probability of
detection, which turned out to be zero in the setting presented; part (c) asked
how to modify the length of one arm of apparatus, in order to maximize the
detection probability. The fifth question was about Young’s experiment with
electrons, and was composed of a quantitative part (a), which is similar to
one of the problems of highest difficulty rating found in the Italian high school
Cutnell-Johnson5 [193], and a conceptual part (b) in which the requirement
5
As many other textbooks, the Italian Cutnell-Johnson includes a system for rating the
relative difficulty of exercise, which uses a variable number of “stars”.

159
 6. Implementation and results

was to discuss the case of a which way measurement. 17 students were present
on the day scheduled for the test. The number of correct answers for each item
is reported in Table 6.9.

Table 6.9: Number of correct answers to the questions of the graded test (total
number of students here is 17).

Compton Uncertainty De Broglie Michelson in- Young’s experiment

effect principle wavelength terferometer with electrons
(a) (b) (a) (b) (a) (b) (c) (a) (b)
17 7 15 5 16 14 7 3 9 9

Although results, as can be seen from the table, were not exceptionally
good, still the test obtained the highest average grade (6.2/10) among all tests
performed by the teacher during the year, which had a similar structure, num-
ber of exercises, and evaluation scheme. Most students reached a “baseline”
level in being able to solve exercises 1 (a), 2 (a) and 3, which are comparable
to the average exercises of their textbook, plus 4 (a) indicating having at least
a general idea of the concept of “possible paths”. Seven students were able to
use the rules of the sum over paths method to obtain the answer that inter-
ference, in the setup proposed, is destructive, but only three were sufficiently
confident with the approach to correctly calculate the answer to item (c). In
the fifth question, part (b) concerned wave particle duality. Nine students
correctly answered that the acquisition of which way information destroys the
interference pattern; among these, 4 proposed no explanation, while other 4
interpreted the fact in terms of reduction of possible paths and / or paths
becoming distinguishable; one student mentioned wave function collapse.

6.3.4.3 Final conceptual test

About twenty days after the end of the sequence, we gave to students a second
test, composed of open questions. The questionnaire is reported in Appendix
A.3.4. In the scheduled day only 14 students out of 18 were present; presum-
ably this happened by chance, as students had been informed that the task
would not have been graded or scored in any way for school purposes. The
three main questions, which were divided in subquestions, probed different
conceptual aspects of quantum physics. The first one was about the uncer-
tainty principle, and required students to criticize a textbook [421] reporting
an explanation of the principle based on the outdated idea of Heisenberg’s
microscope. The second one concerned the role of measurement as seen in the
case of a Mach-Zehnder interferometer in which a which way measurement was
performed. The third one was centered on the two slit experiment with one
electron at a time. Although the third question contained three sub-items,
the first one was not considered in the evaluation as, looking at the student’s

160
6.3. 2015 test with high school students

answers, it was probably formulated in an unclear or misleading way6 . In any

case, in all questions we analyzed the answer as a whole without separating
sub-items, since they very connected and overlapping. In tables 6.10,6.11,6.12
we build knowledge integration rubrics for each of the questions. In Table
6.13, the student-by-student and average grades for the three questions are
reported. Although in Table 6.13 we highlighted the results obtained by

Table 6.10: Rubric for the first question of the conceptual test (uncertainty
principle)

Score Level Description

0 Irrelevant Does not answer the question.

1 Non-normative Agrees with the text, or criticizes it for wrong reasons.
Mentions the idea that uncertainty in is an intrin-
sic property of quantum objects and not due to mea-
2 Partial
surement, makes no other connections or only invalid
ones.
As “partial”, but also discusses at least one of the fol-
lowing: single slit diffraction; uncertainty in classical
and quantum physics; non interaction measurements;
3 Full
uncertainty explaining physical facts; historical evolu-
tion of the concept of quantum uncertainty. Scientific
connection may be incomplete, but essentially valid.
Makes at least one scientifically valid and complete
4 Complex connection with the concepts or examples discussed
in “full”

the students chosen for the final interviews (two of them were not present),
mainly for comparison with the next section, at the time the test was delivered
the choices had already been made and communicated to students, so these
data played no role in the selection. The averages scores for the second (espe-
cially) and third question, which both are 2.4, may be compared to the result
of the 2015 student teachers for the open question exploring similar concepts,
which was 3.3. Although the latter is of course higher, the high school result
still seems encouraging. On the understanding of the uncertainty principle, we
again obtain lower results, with only four students obtaining a “full” or higher
rating. However, a majority of students (9 out of 14) obtains a score of 2
or more, meaning they are able to identify the basic inadequacy in the pre-
sentation of the uncertainty principle as a perturbation due to measurement,
although not always providing valid connections. Reading the answers of those
students who displayed consistent conceptions of the principle, it seems that
the subjects discussed which more efficiently stimulate such models are: non
interaction measurements; the historical development of the interpretation of
6
The expected answer was that, reducing the intensity of light, wave theory would have
predicted the fringes to remain continuous, although dimmer, and not being turned into
individual spots.

161
 6. Implementation and results

Table 6.11: Rubric for the second question of the conceptual test (Mach-
Zehnder and the role of measurement)

Score Level Description

Does not answer the question, or the answer is com-

0 Irrelevant
pletely unrelated to the question.
The concepts introduced in order to answer are wrong
1 Non-normative
or connected in a scientifically invalid way.
Makes a connection between the acquisition of which
way information and the disappearance of interference
2 Partial
effects, but does not try to explain or the explanation
is non-normative.
Explains that which way information reduces the pos-
3 Full sible paths, or that distinguishable processes follow
classical probability rules.
As “full”, but connects the explanation to a complete
analysis of the proposed Mach-Zehnder setup, or gen-
4 Complex
eralizes by making a scientifically valid connection to
a different quantum system.

Table 6.12: Rubric for the third question of the conceptual test (Two slits
experiment with electrons and wave particle duality)

Score Level Description

Does not answer the question, or the answer is com-

0 Irrelevant
pletely unrelated to the question.
The concepts introduced in order to answer are wrong
1 Non-normative
or connected in a scientifically invalid way.
Correctly describes the difference between what is ex-
2 Partial pected on the screen for classical and quantum parti-
cle, but the description is non normative.
Provides a scientifically valid account of how the pat-
3 Full tern on the screen is reproduced using the sum over
paths formalism.
As “full”, but connects the explanation to the case
of the Mach-Zehnder in a complete and scientifically
4 Complex
valid way; or provides a different, scientifically valid,
generalization.

the uncertainty principle; and the derivation of the ground state of the square
well potential solely from the principle.
Besides the quantitative analysis of the ability to connect various scientific
ideas connected with wave particle duality and the uncertainty principle, which
is measured by KI rubrics, it is also worthwhile to analyze some of the hybrid,
fragmented or inconsistent models produced by students, which appear in those
answers obtaining low scores in the rubrics. In the third question on the Merli

162
6.3. 2015 test with high school students

Table 6.13: Student by student and average scores assigned, based on the
codings of Tables 6.10, 6.11 and 6.12, for the final questionnaire. The rows
corresponding to students chosen for the final interviews are colored in cyan.
Rows for students not present the day of the test were left blank to permit a
comparison with Table 6.16 reporting the content of student produced compo-
sitions.

Merli-Missiroli-Pozzi
Mach Zehnder and the
Number Uncertainty principle experiment and wave
role of measurement
particle duality

1
2 2 1 2
3 0 1 2
4
5 4 4 4
6 2 1 1
7 1 2 3
8 0 3 3
9 1 1 0
10 2 3 3
11 3 2 3
12 3 4 3
13 3 2 3
14
15 1 3 0
16 2 3 3
17
18 4 4 4
Average 2.0 2.4 2.4

- Missiroli - Pozzi experiment, only one student gave a completely off-target

answer, which was mainly related to misunderstandings on the behaviour of
classical waves, and also to difficulties in understanding the question. The stu-
dent answered that the electron does not behave as a classical wave, because in
this case it would only produce two superposed diffraction patterns, and does
not behave as a classical particle, because the concept of probability does not
apply to such objects.
The question on the Mach-Zehnder interferometer, although producing on av-
erage the same score in rubrics, highlighted some more conceptual difficulties:
the analyses of two students show clearly deterministic elements. Interestingly,
one of these two students tried to explain the different behaviour of photons
in the two cases shown in the question (see Appendix A.3.4) by hypothesizing
that the in the first case the photon frequency is accurately chosen so that it
is certainly reflected at the second beam splitter, which is revelatory of how

163
 6. Implementation and results

far, in limit cases, students can go in the attempt of maintaining deterministic

conceptions.
In the first question on the uncertainty principle, two students agree with the
analysis reported in the textbook, while one contrasts the views of Heisenberg
with Einstein’s ideas, citing the famous “God does not play dice” quote.

6.3.5 Analysis of cases of appropriation

6.3.5.1 Student by student analysis of argumentative papers
Although many indications for the individuation of focal students represent-
ing possible cases of appropriation came from lesson recordings, the analysis
of student’s written compositions played a primary role for such scope. From
these essays, we could extract data on whether (I) a personal thesis or po-
sition was being developed; or idiosyncratic, signature words expressing an
underlying coherent discourse which was not being made explicit were present;
(II) argumentations proposed, or quotations discussed, were linked to princi-
ples and phenomena firmly grounded in the physical content of the discipline.
Some very partial indications on the first requirement could be guessed already
from the titles which students, as requested, chose for their essay. A few of
the titles were very impersonal and uninspiring (i.e. “From classical to quan-
tum physics”; “A new physics: quantum physics”). Many concentrated on the
themes reported on the guidelines: wave particle duality and the possibility to
provide an intuitive representation of quantum theory (“Quantum physics as
a bridge from undulatory to corpuscular”; “The double nature of the photon
and the representation of reality”). Some contained puns or jokes, which in
most cases are intraducible in English (“A luminous paradox”). Finally, some
expressed very personal views (“Are we the cause of uncertainty?”; “Quantum
theory: physics or metaphysics?”). The titles will be all enumerated later, as
they give a taste of students’ main ideas and thoughts about the subject.
In order to have an approximate measure of indicator (II), we adopted a cod-
ing system inspired by the method of KI rubrics [101, 107] even though, in
the case of an essay, the score assigned must not be considered an assessment
or evaluation, but only an indication. Besides the use in identifying cases of
appropriation, the coding can be seen as somewhat correlated to students’
general confidence with quantum concepts, as expressed in their compositions,
produced about halfway in the sequence development. For uniformity we also
adopted a similar coding for indicator (I), although in this case the assigned
score has nothing to do with knowledge integration. In the following Tables
6.15 and 6.14 the indicators and scores are reported. It may appear by reading
the tables that it is almost impossible for a composition to score a zero on both
indicators. However, this kind of productions are not so uncommon, at least in
the Italian school system, for tasks as the one we proposed; infact there is one
example that we scored in this way among the 18 we had available. Such an
essay may consist in a series of quotations or paraphrases from the proposed

164
6.3. 2015 test with high school students

Table 6.14: Coding system expressing the robustness of grounding in physical

content of student’s argumentation

Score Level Description

No independent disciplinary grounding for the ci-

0 Irrelevant
tation or argument presented
Principles or experiments introduced are misinter-
1 Non normative preted, or have non consequential or very weak
connection to the citation or argument presented.
Overall interpretation of principles and experi-
ments is correct; connection with the citation or
2 Partial
argument presented is superficial or not elabo-
rated.
As “partial”, but at least one principle or experi-
ment among those discussed has a valid and sci-
3 Full
entifically complete connection to the citation or
argument presented.
Two or more experiments and principles among
those discussed have a valid and scientifically com-
4 Complex
plete connection to the citation or argument pre-
sented.

Table 6.15: Coding system for the personal, idiosyncratic content in students’
productions

Score Level Description

Neutral writing style, no identifiable idiosyncratic

0 No expression
keywords, no personal comments, no overall thesis.
No explicit personal position expressed, but id-
1 Implicit expression iosyncratic keywords denote underlying discourse.
At most one incidental personal comment.
More than one personal comment or opinion ex-
Partially explicit expres-
2 pressed, or only one but extended and non-
sion
incidental.
Global personal thesis or overall argument consis-
3 Full explicit expression
tent throughout the paper.

texts, joined by brief, neutral connective sentences (such as “Einstein, instead,

thought that...”), preceded by a short, impersonal, introduction and followed
by a conclusion with similar characteristics.
In table 6.16 we report a list of all students’ papers, with the titles and scores
they were assigned based on the previous tables. The students we chose for
the final interview are in colored rows. As mentioned above, the essay was
not the only factor we considered in this choice; other data were considered,
such as the degree of participation in discussion, and the results of the graded
test, which was performed before the final choice of focal students. As can

165
 6. Implementation and results

Table 6.16: Composition titles and scores assigned based on the codings of
Tables 6.14 and 6.15. The rows corresponding to students chosen for the final
interviews are colored in cyan.

Disciplinary Personal
Number Title
grounding content

How to make concepts of quantum physics comprehen-

1 1 1
sible.
2 Physically7 . 0 2
Quantum mechanics between mathematical models and
3 philosophy
0 2
4 Wave particle duality. 2 3
5 The paradoxes of quantum mechanics. (Chiara) 3 2
6 The language of quantum physics. 1 2
7 Quantum theory: physics or metaphysics? 1 3
8 Are we the cause of uncertainty? 1 3
Overcoming the physics of certainty: “how many” ways
9 1 0
to understand it?8
Quantum physics as a bridge between undulatory and
10 3 1
curpuscular.
11 A luminous paradox. 0 2
12 Quantum physics and wave particle duality. 2 2
13 A new physics: quantum physics 2 1
14 Me and physics. 0 2
15 From classical to quantum physics. 0 0
The double nature of the photon and the representation
16 of reality.
3 2
17 Wave particle duality: law or contradiction? 3 2
18 Randomness in quantum physics. (Cheng) 4 1

be seen from the table, only one student reached level four in the coding for
disciplinary grounding of content. However, his essay was written in a rather
impersonal style. This is the case of Cheng, a student of Chinese ethnicity
and culture, which we will discuss in Section 6.3.5.5. A couple of students
produced quite personal interpretations of the proposed theme, for example
student 7 argued that quantum physics is a modern sort of metaphysics, while
student 8 consistently defended in the whole paper the thesis that the intrin-
sic limitations of our intellect do not allow us to understand quantum physics;
however their argumentations were poorly physically grounded, and based also
on other data, we did not include them in the final interviews, given the strict
constraints in available time. It must be noted that, in all but one case, the

7
This is an intraducible pun. “Physically” in italian is fisicamente which the student
wrote FisicaMente, meaning “PhysicalMind”.
8
This is also a pun since “how many” in Italian is quanti, which also means “quanta”.

166
6.3. 2015 test with high school students

students who scored three points or more in the above table were the same who
actively participated in classroom discussions, a fact that can be seen as con-
firming the idea that knowledge is socially constructed. The only exception,
again, is Cheng, who never spoke during discussion activities.

6.3.5.2 Final interviews

The interview protocol, which is reported in Appendix C, was divided into
three parts:
1. Questions concerning the content and specific concepts. For example, stu-
dents were asked if they were able to reconstruct the various ways in which
the uncertainty principle had been explained to them, and which one they
found more useful for understanding.
2. General questions. For example, students were asked whether the initial
activity on the intuitive and visualizable content of physical theory had
been useful to them.
3. Specific questions to individual students.
While the first two parts of the protocol were the same for all students,
many questions of the third part were specifically aimed at each one of them.
Such questions concerned aspects of their individual experience with the se-
quence; were meant to clarify or highlight things they had said, written or done;
and in some cases intended to probe some of the concepts or idiosyncratic terms
that we had identified as recurring in their essays, and in discussions.

6.3.5.3 Individuation of students’ profiles

At the end of the interviews, we analyzed them using the methodology delin-
eated in Ref [34] in order to produce the profiles of individual students. This
involved the following steps:
1. Find or confirm an internal consistency, consisting of a real “thesis” or of
one or more idiosyncratic ideas recurrent in the interview.
2. Look for, or confirm a possible coherence between the ideas expressed in
the interview and the actions of the student during classroom activities,
and in the composition.
3. As evidence of appropriation, check whether the students’ thesis or idiosyn-
cratic idea meets the criteria enumerated in Section 2.2, i.e. to be grounded
in the discipline, “thick” in an epistemological and meta-cognitive sense, non
incidental, and bearer of social relations.
In this dissertation, we will only analyze the cases of Chiara, who we believe is
a clear example of appropriation, and Cheng, which is the most complex and
faceted one. We leave the analysis of the individual profiles of the other four
students who were interviewed to a future expansion of this work.

167
 6. Implementation and results

6.3.5.4 Chiara: confronting modern physics with logic, and the

search of for an overall logical order.

From the point of view of educational gains alone, Chiara is the most evident
case of success for our learning path. At the beginning of the year, she did not
excel in physics or mathematics; her results were only slightly above average.
This started to change in the course of the year, and the process was highly
accelerated during the development of the sequence. About halfway into it,
Chiara had taken a positive leadership role in physics: she was the one other
students referred to when they were not able solve exercises, or when they
had to discuss conceptual aspects of the sum over paths method, of which she
had acquired a deep-rooted understanding. At the end of the sequence, she
did exceptionally well in both the graded and ungraded tests. Her results in
mathematics also improved in the meantime.
Chiara is a shy person, and initially she did not intervene much in discussions,
although her few remarks were thoughtful and non trivial. As she perceived
that she was excelling in the understanding of both formal and conceptual as-
pects of quantum physics, she took confidence and her participation increased.
At the end of the sequence, she said that for the first time she was consid-
ering the study of physics as one of her possible choices in university. For
her final high school exam, after changing her mind several times, the project
she discussed concerned the Einstein-Bergson controversy on relativity and the
nature of time.
In the initial discussion, Chiara only made two interventions, and in both cases
she insisted on the concept that a physical theory can, only or primarily, be
understood intuitively in terms of another physical theory, which is more fun-
damental or microscopic; she made the example of the magnetic force in a wire,
which can be comprehended starting from Lorentz force on electrons. During
the sequence, at some point in her intellectual trajectory, she developed or
brought to surface a preoccupation concerning the compatibility of modern
science with logic. Once, during a mathematics lesson, she made the remark
that she feared that science was progressively abandoning logic. At that time,
her idea for her mini-thesis project for the final exam was to discuss the crisis of
the logicist foundational program in mathematics, treating Frege’s Grundlagen
and Russel’s paradox, a subject that is not part of the high school curriculum,
and which she had discovered on her own. Also, in some of her interventions in
class she wanted to discuss aspects that she perceived as paradoxical, such as
the fact of how can, in the Zhou-Wang-Mandel experiment, one photon appear
to “know” that the other one has been detected.
Chiara’s argumentative paper was titled “The paradoxes of quantum mechan-
ics.” The paper is rather short, perhaps testifying an interlocutory phase in
her development. In the essay, she discusses the “great paradoxes” of quantum
physics, which “under many aspects appears contradictory”, such as admitting
the realization of possibilities that “apparently seem mutually exclusive”. She
strongly highlights Lévy-Leblond’s idea that a new specific concept must be

168
6.3. 2015 test with high school students

created, giving to it a new name; and indeed she adheres to the use of the
term “quantum object” throughout the paper, as she will do also in the final
interview. She also discusses Einstein’s radical criticism to the theory, but con-
cludes that “Einstein’s resistances were overcome thanks to the great predictive
power that formulations of quantum mechanics have shown in successive ex-
periments.” We highlighted the term “formulations of”, which appears to be
redundant in the above sentence, since may allude to the possibility that dif-
ferent formulations of quantum physics have a different content concerning
its paradoxical aspects, which is in fact true. The paper also contains a par-
ticularly effective account of non epistemic uncertainty in quantum physics:
“Uncertainty is an intrinsic property of objects of the quantum world, and
consists in the fact that only following the act of measurement a real value
can be obtained, but until the measurement is not performed, the quantum
object remains in a state which is objectively uncertain, and only describes a
“potentiality” of the object itself.”
At this point, Chiara’s underlying epistemological discourse seemed to have
been delineated in its general lines: it concerned the compatibility of quantum
physics, and modern science in general, with logic; and the possibility of for-
mulating a theory in such a way that paradoxes can be avoided, or reduced.
It was with this idea in mind that we designed the individual questions for her
in the final interview.
In the initial part of the interview, Chiara gives an account of several con-
cepts of quantum physics, such as the uncertainty principle, the De Broglie
wavelength, and recapitulates the elements of the sum over paths approach.
Her choice of language and vocabulary is always very careful, she sticks to the
term “quantum object”, rather than particle, and she displays a non superficial
acquisition of the concepts studied. As expected, she uses the term “paradox”
or “paradoxical” repeatedly in the interview (7 times). The first of such occur-
rences the is when, after discussing the uncertainty principle, the interviewer
asks “what adjective would you use to describe this property of the quantum
object, for example clouded, veiled...” she answers “probably... paradoxical.”
Another idiosyncratic term in the interview is “understand” or “understand-
ing”, which although being a very common term in this kind of interview, she
repeats a very large number of times (11).
In the course of the interview, it becomes progressively clearer that identifying
the paradoxical character of certain aspects of quantum physics does not nec-
essarily prevent Chiara from accepting them. This can be exemplified by the
following passage of the interview, which occurred after she provided a very
clear account of how acquired information reduces the possible paths, negating
interference:
Interviewer: ”Ok, we see that you understood the concept that information
reduces the possible paths very well, but we wanted to know what you think
about the matter.”
Chiara: “Yes well, in fact when the teacher said... at the principle I thought I

169
 6. Implementation and results

had misunderstood, but then the teacher said that I was right, it appears that
the photon knows that it is being observed... it was rather paradoxical.”
Interviewer: “And how did the teacher convince you of this?”
Chiara: “Uhm, he convinced me.”
Interviewer: ”Yes, but how? Just because you trust him?”
Chiara: ”No, always for the same reason... because studying, making exper-
iments, it appeared evident that things went this way, also from the mathe-
matical model.”
Interviewer: “And does it not disturb you?”
Chiara: ”Yes it does... or maybe not, it makes things more interesting. It is
good to know how things work, but it is also good to know that they can work
differently from what we expect.”
Interviewer: ”It fascinates you.”
Chiara: “Exactly, it fascinates me.”
Concerning the uncertainty principle, she expresses similar views. At first,
she was disturbed, thinking it was paradoxical. But then, when she realized
that experiments inescapably demonstrated that the principle actually occurs
in nature, she started to accept and even appreciate it, realizing it has an ex-
planatory value. Reporting Chiara’s own words:
Interviewer: “Do you think the uncertainty principle casts doubts on the
possibility of knowing or understanding the world?”
Chiara: “Probably the opposite. Rather than casting doubts on it, it makes
you understand that the previous model did not work perfectly in describing
reality, so on the contrary it makes us understand more.”
Interviewer: “So it gives us a better model to describe the world.”
Chiara: “Better, and which makes us understand that things are not so sim-
ple as we thought, trajectories, forces...”
Interviewer: “So this aspect of quantum mechanics is not so astonishing.”
Chiara: “It is astonishing, but in a positive sense.”
Immediately after, the interviewer asks Chiara if there is something that she
really finds disturbing in quantum physics. The resulting short passage is also
important for completing the delineation of Chiara’s epistemological profile
Chiara: “I think... probabilities. Intrinsic probabilities. The fact that the
photon can go one way or the other and you are not able to determine... you
cannot understand why it takes one route or another.”
Interviewer: “Well, if I had to say what astonishes me in quantum physics,
it is not that the photon can go with a certain probability through one way or
the other but..”
Chiara: “No, not being able to understand where it may go.”
Although it is not easy to say exactly what Chiara has in mind here, taken as a
whole the previous passages provide a clearer picture: her preoccupation with
paradoxes and the perceived logical contradictions of quantum theory is not
of an abstract nature; she is concerned that such paradoxes are an obstacle to
understanding nature, and she even appreciates them, as long as they provide

170
6.3. 2015 test with high school students

a better understanding. As far as a logical order goes, this concerns Chiara

because she views logic as the primary instrument for understanding, but she
is even open to the possibility that, for quantum mechanics or other theories
in modern physics, a different logic structure is required in order to reach un-
derstanding. This is revealed by the following passage of the interview, which
belongs to the final part, when individual questions are asked:
Interviewer: “The teacher said that you are preoccupied that science in the
twentieth century is abandoning logic. Would you like to tell us more about
this concern?”
Chiara: “No, but I only meant classical logic.”
The interviewers did not immediately catch the hint here, but it would have
been interesting to know what other kinds of logic Chiara refers to, since even
classical logic is treated very superficially in the high school curriculum. How-
ever, we have some indications on this a bit later:
Interviewer: “In the final test, you wrote that results obtained are surpris-
ing, but “not incomprehensible, because they can be demonstrated visually
and mathematically by considering certain characteristics as intrinsically true”.
Thus, in the end, did quantum physics satisfy you requirements of logical
rigour?”
Chiara: “No, because of its paradoxical aspect... well it depends, my idea
was about classical logic, but it depends on how you interpret the concept of
“logic”. Originally, I thought that logic was simply common sense, what we
are used to think. Then I reflected upon it, and I said to myself: ”Logic is not
simply common sense.” So I asked to myself: “What is logic?” and I started
from Aristotle, deduction rules... then I learned about mathematical logic...”
Interviewer: “But according to you, what principles of classical formal logic
are contradicted by quantum physics?”
Chiara: “Certainly the law of excluded middle.”
Is Chiara aware of the existence of fuzzy logics? We do not know. But clearly
the interview shows us that her idiosyncratic idea of confronting quantum
physics with logic is much more thick than we could imagine before conduct-
ing it. In fact, it has deep roots in meta-cognitive (how do I understand?
According to what rules? Could I understand according to different rules?)
and epistemological (how is new knowledge about the world acquired?) di-
mensions.
Most of the other operative markers of appropriation introduced in [34] are
also easily confirmed. The idea is certainly authentic, in the sense that it re-
flects her inner epistemological conceptions, and is expressed in a language and
using terms that are not borrowed from the teacher or the textbook. Actually,
it appears that in the course of the year her idea has evolved through a signif-
icant amount of personal research. As we have shown in the beginning of this
section, it is also clearly non incidental. It is disciplinarly grounded, not only
in the sense that, given Chiara’s exceptional results in quantum physics, the
paradoxes she identifies are always discussed fully respecting the rules of the

171
 6. Implementation and results

discipline, but also because her general attitude towards apparently contradic-
tory aspects is firmly grounded in the founding norms of physics. In fact, in
her search for a logical order, she makes no compromise with the principle that
the results of experiments must be accepted. Nature must be understood as
it is, and its strangeness even fascinates Chiara, since it transforms paradoxes
in tools for understanding. In the end, if a contradiction persists, then it is
classical logic and our usual way of thinking which must cede the way, as she
hints to in the final part of the interview.
Is Chiara’s idiosyncratic idea carrier of social relations? This is perhaps the
most difficult to verify, among the five operational markers of appropriation.
Chiara was not viewed in the class as “the logician” or “one obsessed with para-
doxes”; at most, some of her friends would have described her as complicated.
Certainly she can be depicted as a person wishing to understand. Probably,
the best way to interpret this marker in this case, which is clearly one of appro-
priation, is that her urge for understanding and reconstructing a logical order
led her to become the recognized reference for the rest of the class in terms of
disciplinary content [34].

6.3.5.5 Cheng: a student on his way to appropriation

At the beginning of the year, before Chiara’s positive evolution, Cheng was
considered by classmates the only student of the class to be really good at
mathematics and physics. His results in these disciplines had been exception-
ally good throughout high school, and indeed continued to be so during the
fifth year.
Cheng is a first generation immigrant, arrived in Italy only at the beginning
of high school. While he speaks a good Italian, and understands and writes
it almost perfectly, his cultural background is fully Chinese. It is also likely,
although we do not know for sure, that he is, in fact, from rural China. One
indication in this sense is the confession he made to several teachers that,
notwithstanding his desire to do so, he would probably not be allowed by
his parents to attend university, being instead destined to work in the family
restaurant. This is rather typical of immigrant families of rural Chinese cul-
ture, as is the fact that his parents never in five years showed up at encounters
with the class teachers. Cultural aspects cannot be ignored when considering
Cheng’s style of social interaction. As an example, when asked by the teacher
why he was always silent during discussions, he answered that, in his culture,
to intervene in class means to “show off”, and is not well regarded; and that
he did not think he was authoritative enough to “show off”. An element of
personal shyness has also probably to be added to the picture.
Initially, Cheng had difficulties with the quantum physics sequence. We know
this not only because he says so in the final interview, but also because at first
he could not solve the additional exercises on the sum over paths approach that
were given as home assignment. Being used to study alone, based primarily
on the textbook, it is possible that he could not find enough information on

172
6.3. 2015 test with high school students

how to deal with exercises from the slides of the sequence alone, and he was
confused. However, after enough problems were dealt with in class, he quickly
catched up with the new method, and at the end of the sequence his results
were exceptionally good on both the graded and not graded tests.
Since Cheng was, literally, always silent during lessons and discussions, the
only data we have on him are his written productions, and the final interview.
His argumentative paper was a longer than average and well documented essay,
titled “Randomness in quantum physics”, which the Italian Language teacher
rewarded with a 8/10 mark. In it, he traces back some of the most important
developments of quantum physics, in a rough but essentially correct histori-
cal order, producing in all cases sensible, and sometimes detailed, accounts.
Initially he discusses the photoelectric and Compton effect, and the discov-
ery of the wave behaviour of electrons and neutrons. The central part of his
composition is devoted to presenting the various perspectives on wave parti-
cle duality, and he introduces the ideas of Schrödinger, Heisenberg, Einstein,
and Lévy-Leblond. Next, he discusses some more recent developments, such
as the Zhou-Wang-Mandel experiment, which he describes in detail, and the
phenomenon of entanglement, which is the result of personal research, as it was
not treated in class. The final sentence of the text is also the only personal
comment present in it: “It seems absurd that the world is governed by chance,
but quantum physics suggests precisely this.”
The only idiosyncratic feature that could be identified in Cheng’s composition
was a linguistic habit of presenting the ideas of scientists of the past using the
terms “explains” or “explained” (used four times) or alternatively “said” (used
twice) rather than the more common “thought” or “wrote”. What these verbs
may have in common is that they are more easily referred to verbal utterances.
This is also consistent with his final conceptual test, in which, in the answer
to the first question, he concludes that “The textbook is dated because follows
what Heisenberg said”. In the same test, realizing that the answer to question
3 requires some of the concepts already used for question 2, he writes “As I
have explained before...”
A second characteristic feature of Cheng, which does not appear in the com-
position, is the importance he attributes to mathematics in the intuitive un-
derstanding of physics. For example, in the initial questionnaire he writes:
“Images can help you understand, while the mathematical model simplifies
everything. If we know how it works, it makes us remember everything at a
glance.”
And also:
“In mechanics, for me the central role was played by the mathematical model,
because it is very concrete, so I can imagine, while for more complicated sub-
jects I used the other two instruments [images and analogies] more.”
And in the final questionnaire, question 2 (b), explaining why the results of a
which way measurement may be surprising, but not incomprehensible:
”It is surprising because it does not follow the classical probability rule, but is

173
 6. Implementation and results

not incomprehensible, because it follows the quantum probability rule. So it

is surprising, but only because it is computed in a different way.”
Thus before the interview we had identified two weak, apparently disconnected,
idiosyncratic features: a habit of using the verbs “explains” or “says” to report
the ideas of scientists of the past, which could have been (and which we still
aren’t sure isn’t, at least in part) related to the acquisition of Italian as a sec-
ond language by a native Chinese speaker; and a tendency to give more weight
to the role of the mathematical model in intuitive understanding of a physical
theory. In the individual questions of the third part of the questionnaire, we
only attempted to probe explicitly the second of these two ideas.
Cheng’s interview progresses with some difficulty, partly also because his Ital-
ian is good but not perfect. It is made of short answers, whose content is
sometimes perplexing. In the beginning, he reports that conceptual aspects of
quantum physics were easy for him, while he had initially difficulties with the
exercises related to the sum over paths approach. Among the properties of the
quantum object, he mentions wave-particle duality. The interviewer here tries
to delve deeper in his understanding of wave particle duality, and Cheng spec-
ifies that its reason can be found in the acquisition of which way information,
which reduces the possible paths, negating interference. The honest impression
is that he says so in order to inform the interviewer that he knows the correct
answer; in fact, a few seconds later, the interview goes on as follows:
Interviewer: “... but then, considering Lévy-Leblond’s example of the platy-
pus, is there actually a duality, or there is none?”
Cheng: “Eh, because still a solution hasn’t been found; and until a solution
is found, duality is used.”
Interviewer: “So in your opinion a solution hasn’t been found.”
Cheng: “Not yet.”
Interviewer: “So you are not convinced by the idea of a quantum objects,
which is neither wave nor particle (...) You believe a better explanation exists.”
Cheng: “I think it exists, but hasn’t been found yet.”
Interviewer: “So your are not convinced.”
Cheng: “I am convinced that as long as we don’t find it we cannot use it.”
A few lines later, Cheng says even more clearly that he does not believe quan-
tum physics to be the final answer:
Cheng: “I would like to discover why it is that way.”
Interviewer: “Is it my impression or there is something that you do not ac-
cept. ”
Cheng: “Exactly. I would like to know more about reality.”
Interviewer: “So you don’t accept it. Sooner or later it will be discovered.”
Cheng: “Yes. Exactly.”
The interviewers continue probing these concepts for a while, using different
examples, and Cheng consistently repeats that he believes that a better an-
swer will be found, and the he accepts quantum physics as provisional, as a
mathematical model that works. Concerning the uncertainty principle, how-

174
6.3. 2015 test with high school students

ever he carefully specifies that he does not believe it to be a consequence of a

perturbation due to measurement:
Cheng: “I believe objects to have a definite position and momentum. There
is something that escapes our understanding. But it is not that uncertainty is
due to measurement. It due to some other reason. Something which we still
don’t know.”
In fact, “something which we still don’t know” and equivalent expressions ap-
pear almost obsessively in this part of the interview. Cheng, however does
not appear to build misleading mental images, or try to invent new theories.
When talking about the two slits experiment, he again declares that “there is
something we don’t know yet”; but then, when he is asked about his mental
image:
Interviewer: “Do you have a mental image of what the electron does in the
two slits experiment?”
Cheng: “It passes through bot slits... but it is indivisible.”
Interviewer: “But it passes through both slits.”
Cheng: “Yes”
Interviewer: “So in the end you are visualizing Feynman’s model.”
Cheng: “Yes”
Interviewer: “What do you find strange in this experiment?”
Cheng: “That it is able to pass through both slits. Without splitting.”
After some more lines, the interviewer asks:
Interviewer: “How would you convince your friends that quantum mechanics
is strange, but still we can understand it? ”
Cheng: “It is because it works with experiments, we can compute many things
with accuracy.”
The exchange that comes immediately after has been perhaps the most im-
portant interpretative key to Cheng’s epistemological conceptions. The second
part of the interview has just began, and the interviewer is asking about the
discussion activities based on texts from great scientists of the past:
Interviewer: “...those epistemological reflections have been useful for you, in
your learning path? ”
Cheng: “Yes, because we can know the history and the origin of quantum
mechanics, read the various interpretations of scientists, and we can think in
a similar way to them.”
Interviewer: “Er... what did you just say? Why are the interpretations of
scientists so important to you?”
Cheng: “So we can compare our ideas to theirs, and see whether our ideas
are right or wrong.”
Interviewer: “And what did you conclude from such comparison?”
Cheng: “What I said before, there must be a mechanics which is even more
accurate, but still hasn’t been found. ”
Interviewer: “Yes, but who do you refer to, among these scientists, for draw-
ing such conclusion? Someone in particular?”

175
 6. Implementation and results

Cheng: “No, none in particular. It’s a synthesis.”

When asked whether the path had stimulated discussions with the rest of the
class, he says “For me, not much. I only discussed exercises, because they were
difficult.” However, he also says that the very interested by the sequence. At
this point the interviewer asks him a very important question, namely if he
ever discussed with classmates his doubts on quantum physics, and he answers,
appearing mildly amused by the question, that he didn’t.
In the final part of the interview, again the question of wave particle duality
comes out, and again Cheng takes up the attitude of “suspension of disbelief”
and without difficulty discusses it from the point of view of Feynman paths.
Then, the interview turns on the role of mathematical models:
Interviewer: “So for you, the most important feature that makes a phe-
nomenon comprehensible, is to have a coherent mathematical model? ”
Cheng: “For me, yes”
After some lines:
Interviewer: “Is Nature a book written in mathematical characters? ”
Cheng: “Yes.”
Finally, after some more lines he is asked:
Interviewer: “So you believe Newton’s formula for gravitation exists some-
where, and we just discover it. ”
Cheng: “It exists, in the sense that it’s intrinsic. But it’s not mathematics.
We mathematise it.”
Let us summarize all we know about Cheng’s idiosyncratic ideas expressed
in the interview, and the connections to his epistemological conception. The
recurring phrase is evidently “There is something that we don’t know yet” and
its variations, which appears at least 10 times, in discussing different physical
concepts, in the first and second parts of the interview. Such phrase is con-
nected to an idea which is very clearly expressed in the interview, and in the
interview only, that a better theory than quantum mechanics will someday be
found, when we will know more. According to Cheng, he came to this con-
clusion by confronting his ideas with those of the great scientists of the past,
trying to think in a similar way to them, and realizing that a synthesis could
not be found. But there is more: since according to him the most important
factor that makes a theory comprehensible is its mathematical model, not only
he can happily live with quantum theory in the meantime, since “a theory that
does not yet exist cannot be used”, but he can even honestly declare that quan-
tum physics is understandable, and he visualizes it using Feynman’s model.
All this is perhaps a bit disheartening from our point of view as educators,
but we cannot deny that Cheng’s idea is very personal and also thick, it he
sense that, according to Cheng himself, it has a very strong connection with
an historical-epistemological dimension which was solicited by the activities
we proposed.
Is it grounded in the discipline? It would be very difficult to argue that the
idea that “there is something we don’t know yet” explicitly violates the disci-

176
6.3. 2015 test with high school students

plinary norms of physics. As we mentioned in discussing the interview, the idea

does not lead Cheng to produce misleading or invalid images of physical phe-
nomena, since, for the time being, he provisionally accepts the mathematical
model of Feynman paths, and from this he derives his intuitive understanding.
However, it is also true for the same reason that the idea does not help him
in any way in endowing quantum physics with meaning, or in coordinating its
content. It is simply an underlying mental reserve, which he has to suspend,
taking a different perspective, when actually discussing physical phenomena.
Is the idea non incidental? Reading it again after the interview, Cheng’s final
comment in the argumentative paper, “It seems absurd that the world is gov-
erned by chance, but quantum physics suggests precisely this.” can be seen as
one, very watered down appearance of the idea “There is something we don’t
know yet”. Apart from this, assuming that such idea predates the interview,
it has been carefully hidden from the classmates and the teacher. Probably
not primarily in fear of a negative mark, but because Cheng’s inherited social
norms prevented him from expressing an idea which opposed the one of the
teacher. All of his written productions were honestly indistinguishable from
those of someone who had interiorly accepted the quantum world view. The
only possible other, very weak, hint we had, not of the idea itself, but of its
epistemological grounding, was a visible tendency to represent physicists of the
past as if they were speaking and explaining things, an idiosyncrasy that may
be due also to linguistic interference.
Is the idea carrier of social relations? Since it was declaredly hidden from
classmates, probably the only reasonable answer here is no. The fact itself
that it was hidden is probably due to social and socio-cultural issues, such as
the fear of a negative reaction and a feeling of inappropriateness, but this is a
different matter altogether.
Overall, we are prone to interpret the case of Cheng as one of non appropri-
ation, or incomplete appropriation. This is perhaps a bit disturbing, for his
disciplinary results in tests were excellent, second only to those of Chiara, and
he also demonstrated in the argumentative paper a very strong hold on the
historical and experimental development of quantum physics. But, we feel that
we could declare Cheng as a case of complete appropriation only by unaccept-
ably stretching the meaning of at least two, perhaps three of the operational
markers in Ref. [34]. More importantly, we feel that those markers are actually
meaningful, since Cheng at heart has not really accepted quantum physics as a
fundamental theory; and his results could probably have been even better if his
doubts and mental reserves had been shared with the teacher and classmates,
and explicitly discussed in class, and among students.
However, adopting a more optimistic point of view it can be certainly said of
Cheng that he represents the case of “a student on his way for appropriation”.
In the language of coordination class theory [422], Cheng showed lack of span
and articulated alignment, but he started to develop a personal view, as well
as showing outstanding understanding of the contents. Another consideration

177
 6. Implementation and results

we may add is that, in this case, complex cultural and linguistic factors could
have interacted with the processes favoring appropriation, which may deserve
further study.

178
Chapter 7
Conclusions and future
perspectives

This final chapter is divided in two sections. In Sec. 7.1 we will reconsider, one
by one, the research questions which were formulated in the introduction (Sec.
1.2), proposing tentative answers which emerge from our work. Despite several
positive and encouraging results, in many cases the answers are incomplete, and
future work necessary to improve such answer will be discussed. In the brief
Section 7.2, we discuss future plans for the expansion of this dissertation work,
not specifically related to the research questions, including possible expansions
of the disciplinary content covered.

7.1 Re-evaluation of research questions

Q1: Can we transfer the way in which the Feynman approach is used by scien-
tists, particularly in the field of quantum optics, to the high school context,
making it become a language to discuss experimental setups and concep-
tual issues, first and foremost those concerning wave particle duality?

On this research question we obtained perhaps the most convincing and impor-
tant positive results. According to data collected in all three our experimenta-
tions, two with student teachers, and one with high school students, the use of
the Feynman approach really is effective in leading students to build consistent
mental models of wave-particle duality. Depending on how the questions were
asked, 80 to 100% of student teachers (see results of Sec. 6.2.3.5), and 50 to
65% of high school students, could provide a scientifically acceptable depiction
of wave particle duality, based either on the reduction of possible paths due to
acquired information, or to the fact that, when alternative processes become
distinguishable, the classical probability rules are used. Furthermore, even for
those students that (mainly in the high school context) did not provide a satis-
fying explanation for phenomena connected to wave particle duality, we rarely

179
 7. Conclusions and future perspectives

observed the appearance of hybrid conceptions, but mostly of phenomenolog-

ical descriptions lacking a connection with normative ideas. In particular, the
reported difficulty that, when instructed using the sum over paths approach,
students may believe that in the end the quantum object follows one of the pos-
sible paths appeared in only one student in the analysis of the Mach-Zehnder
interferometer with intermediate which way measurement. Thus, according
to our results, which confirm and extend those presented in Ref. [145], the
use of the Feynman perspective appears appropriate to overcome students’
difficulties in building mental models of wave particle duality. Many students
and most student teachers in our trials also appeared comfortable in interpret-
ing, in open response item, quantum optics interferometry setups, such as the
single photon Mach-Zehnder, the Zhou-Wang-Mandel and Hong-Ou-Mandel
experiments, in terms of the Feynman’s perspective; and we did not observe
in any significant way the misinterpretation of paths as trajectories (see also
the discussion of Q4, Q5 and Q6). Notwithstanding some initial concerns, the
Zhou-Wang-Mandel experiment was generally understood in its basic meaning
by high school students, played a crucial role in demonstrating the generalized
idea of wave particle duality as interference between indistinguishable pro-
cesses, and non-interference between distinguishable ones [18, 423].
Concerning the uncertainty principle, our results are less clear-cut. In our first
test with student teachers, we obtained an improvement in completely accept-
able depictions of Heisenberg’s relations from 0% in the pre-test to about 42%
in the post-test. In the 2015 test with student teachers, due to various ex-
ternal constraints, we did not perform a test on this specific issue; but in the
parallel experimentation in high school, 64% of students criticized the Heisen-
berg microscope view of uncertainty for reasons compatible with the current
understanding of the principle, although only 29% could provide independent
arguments sustaining such criticism. This discrepancy constitutes, to be com-
pletely honest, a possible signal of non-appropriation of the subject by some
students, as it is possible that many of the 64 − 29 = 35% of them who wrote
that the uncertainty principle has to be interpreted as an intrinsic property
of the state, without providing any further arguments in favor of such claim,
were repeating an argument borrowed from the authority of the teacher. At
the present state of research, our provisional conclusions on this points are that
the Feynman perspective is not less effective than other teaching approaches
in leading student to a full comprehension of the uncertainty principle, but it
still must be complemented with other inputs, such as experimental results,
physical situations and problems in which it has an explanatory value, such as
the tunnel effect and the determination of the ground state energy for confin-
ing potentials; or discussion of epistemological controversies. In particular, as
reported also in Ref. [199], the association between the uncertainty principle
and the ground state energy for a particle in a box seems effective in favoring
the construction of consistent models by students.

Q2: Can we, also through an in-depth investigation on the disciplinary content,

180
7.1. Re-evaluation of research questions

perform a suitable educational translation of the sum over paths represen-

tation of the Green function, obtaining a consistent method for dealing
with time independent problems from a sum over paths perspective?
Our work on the sum over paths representation of the Green function resulted
in the consolidation of a teaching method for introductory time independent
one dimensional problems which, as far as we know, has never been experi-
mented before [340]. Subjective impressions from teaching experience in the
first tests were positive, but certainly most of the work on what are the difficul-
ties and mental models of students in learning the subject with this approach,
which is very different from the standard one, is still to be done. Our hypothe-
sis is that difficulties related to energy loss in tunneling, or incorrect depictions
of wave functions as bi-dimensional objects in one-dimensional square well may
be reduced; on the other hands, it is possible that different difficulties appear.
In presenting the atomic model, our approach was to start with the problem
of quantization on a ring, which we solved using the same method as for other
one-dimensional problems, then derive a model equivalent to the Bohr one,
and discuss in depth its limitations, with a qualitative introduction to the or-
bital model, and finally switching to the energy level representation. In this
area, our presentation certainly must be improved, trying to build a stronger
connection between the sum over paths approach and the orbital model. Al-
though the mathematical difficulties in this case are formidable, they could be
at least partially overcome through the use of simulations.
In the test with ST, we obtained positive results on a common difficulty related
to the energies of emitted and absorbed photons in atomic spectra; however,
students’ models and depictions of the atom have to be more carefully moni-
tored in the future, to determine whether the use of Feynman’s approach could
be of help or hindrance in preventing the persistence of classical, deterministic
views.
Q3: Is the sum over paths formulation at least as efficient as other approaches
in enabling students to find quantitative answers for typical introductory
problems and exercises in quantum physics?
Our attempt at integrating instruction using an innovative, research-based ap-
proach with a textbook using a traditional method can be considered success-
ful. Students did not have more difficulties than in other subjects in solving
textbook exercises, and their results in the final graded test were on aver-
age slightly better than in previous tests. It must be remembered again that
textbook exercises often require little more than the application of a formula;
however, students never complained that the teaching approach was different
from the textbook one, and many appreciated the additional insight provided
by the lesson slides based on the Feynman perspective.
On the other hand, high school students had several difficulties with the ad-
ditional exercises, produced by us, concerning the sum over paths approach,
and many did not even attempt at solving them at home, waiting for them to

181
 7. Conclusions and future perspectives

be discussed during lessons. In many cases, however, those who did succeed in
solving non standard exercises on their own, retained the impression of having
reached important achievements, and obtained significant educational gains in
terms of self-esteem, confidence with the new method and its meaning, and
appropriation. Some student teachers had similar difficulties, but in this case
the possibility, which was not available in the high school setting, of coopera-
tively discussing the exercises online, among them and with their tutor, helped
to overcome such problems.
Q4: Can we, by properly organizing the sequential presentation of the content
in a non-historical order which takes into account the possible generation
of inconsistencies in students’ models, prevent the formation of hybrid,
synthetic conceptions?

Q5: Is the introduction, at precisely selected points in the sequence, of modern

experiments, which sharply highlight the non classical features of quantum
theory, an effective strategy for contributing to achieve the above goal?
We consider these two research questions together, as they are very tightly
connected. In our work so far, most of our concerns on countering students’
formation of synthetic conceptions were focused on wave-particle duality (see
Sec. 3.1.1.1 in the review chapter); the uncertainty principle (Sec. 3.1.1.3);
and the specific deterministic conceptions which, according to some authors
[15], may result from the adoption of the sum over paths approach. In general,
our initial planning of the sequential development content was satisfactory,
and we did not feel the need of substantially modifying it, except for a more
thorough discussion of the uncertainty principle, between the 2014 and 2015
versions of the sequence. As discussed above, we obtained very good results
on students’ models of wave-particle duality and we did not observe, perhaps
with a couple of exceptions in high school, students remaining attached to the
concept of trajectory. More importantly, we had evidence that our order of
presentation of the material, and the experiments we chose, were relevant for
preventing the formation of hybrid conceptions. For example, in the 2015 test
with student teachers, before showing the video for the two slit interference
with one photon at a time, we did a sort of predict-observe-explain experiment,
by directly asking students what were their hypotheses on the nature of light
interference, given the existence of photons demonstrated by the photoelectric
effect. The prevailing theory which came out was that interference was due to
photon-photon interaction. When the video was then shown, ST immediately
realized that their ingenuous theory had been disproved. In the high school
experimentation, during open discussions and interviews, students often rea-
soned that apparently, the simplest explanation for interference would be that
the photon splits in two at the slits, but they added that this had bee been
disproved by the Grangier experiment. The Zhou-Wang-Mandel experiment
decisively convinced several students that the reduction of possible paths due
to a measurement is not due to a physical disturbance on the observed system;

182
7.1. Re-evaluation of research questions

and by extension, although this was not one of its primary aims, was one of
the factors which convinced them that the uncertainty principle also is not
due to such a perturbation. The comparative analysis of the Mach-Zehnder
with removable arms; and of a two slits experiment in which one slit can be
closed, had a profound effect on students, and almost none of them, in the end,
believed that the photon, in Feynman’s model, takes only one of the possible
paths 1
In the relatively more advanced part of the sequence, dealing with time in-
dependent one-dimensional problems, more iterations of the design-testing-
evaluation cycle are needed to determine an optimal sequential development
of the content, and of the possible introduction of experimental evidence, in
the perspective of helping students build consistent mental models.
Q6: What are the main criteria to be respected for producing interactive simu-
lations using the sum over paths approach which are effective in sustaining
students’ understanding, but at the same time do not mislead them into
attributing classical properties to the quantum object?
The issue of visualization has been a major concern in our work. Simulations
giving to students the wrong impression, even with subtle hints, that quan-
tum objects retained classical features, such as trajectories, could be a decisive
factor in leading them to produce inconsistent conceptions . Our efforts, in
general, have been directed to making clear that our simulations should not
be intended as a visualization of reality, but of a mathematical model which,
only in the simplest cases, can be set into a one-to-one correspondence with
three-dimensional space. In this respect, an important guideline we followed
was to always stick to the source-to-detector philosophy, in order to focus stu-
dents’ attention on the emission and detection events and the paths between
them, rather than on the quantum object itself, which was never directly repre-
sented. For the same reason, we did not introduce any visual difference in the
representation of paths related to whether a photon or a massive particle was
involved: the model is in fact in most cases identical, and the only difference is
whether a mass parameter to be varied is introduced. An exception is where,
as in the case of the Mach-Zehnder interferometer, the setup includes beam
splitters. Here, after several trials, we found most productive to represent them
as simple triangular prisms (although the simplest beam splitters are actually
double triangular prisms, glued on their large sides using a dielectric coating)
because it turned out that using this picture made much easier for students
to compute the phase shifts between paths associated to a different number of
external and internal reflections.
We tried to maintain in all simulations a consistent general representation
scheme, showing in one of the windows a sketch of the experimental setup,
and in the second one the details of the calculations of amplitudes and proba-
bility using vectors. In the case of stationary systems, we tried to promote the
1
As we will briefly discuss in the answer to the final research questions, however, some
students believed that a different, undiscovered deterministic model should exist.

183
 7. Conclusions and future perspectives

association between higher energy and shorter (local) de Broglie wavelength

by including such quantity in the visualization, as the energy is varied; and
to never represent amplitudes and energies on the same axis system, in order
to reduce the possibility of students interpreting higher amplitudes as higher
energies. However, in this case we feel that much work is still to do in de-
ciding what is the most productive visualization strategy for simulations of
time-independent problems. This aspect is of course correlated to our plans
of researching in more detail what models students produce of tunneling and
bound systems when they are treated using the sum over paths representation.
Concerning the usefulness of simulations in learning, we consistently had pos-
itive feedback, both from student teachers and in the high school setting. In
some cases, it was also confirmed that the use of GeoGebra, which is widely
used and known by mathematics and physics teachers, can encourage them to
modify existing simulations, and produce their own ones. For example, during
the 2015 test with student teachers, two of them who were particularly expert
with the software produced original simulations representing some of the ex-
ercises which had been assigned.
Besides the work on improving existing simulations, much work is still to do
in producing new ones. For example, experiments such as the Zhou-Wang-
Mandel and Hong-Ou-Mandel ones, which play a very important role in our
sequence, still have not been implemented. Ideally, such simulations should not
only show what is happening in three-dimensional space, something that for
example the QuantumLab interactive representation of the Hong-Ou-Mandel
setup does quite effectively, but also show, with a representation similar to
Feynman diagrams, how the experiment can be interpreted in terms of am-
plitudes of alternative processes. It is possible that, notwithstanding all its
attractive features, GeoGebra in the end is not sufficiently powerful for com-
plex visualization tasks, and that we will have to switch to a different platform
for providing an effective visualization of more complex experiments.
Q7: Can we sustain the work of teachers by designing low cost experiments
which are meaningful for the understanding of quantum physics?
In the current version of or sequence, experimental activities accessible to prac-
tically every high school laboratory play a very important role. The pivotal
element in all these activities was a simple methodology employing digital
photography and image processing techniques which we used to realize a home
made spectrometer based on the use of either transmission or reflection diffrac-
tion gratings, suitable to be employed for inquiry and exploration activities
with groups of students (see Appendix D).
This simple equipment allowed students to measure the wavelength of visible
lines of Balmer series from the hydrogen atomic spectrum, and estimating the
value of Rydberg’s constant with an error difference of few tenths percent [59].
We also employed it to improve the accuracy and experimental value of the
well know activity consisting in the measurement of the Planck constant using
LEDs of various colours. At the end of the sequence, we asked both to student

184
7.1. Re-evaluation of research questions

teachers (in the written activity reports) and to those high school students
who were individually interviewed their comments about their experience and
impressions on these experimental activities. Comments were overwhelmingly
positive, although some conditional remarks were expressed: for example, some
high school students complained having performed the experimental activities
without a sufficient understanding of their theoretical bases, and thus not being
always able to understand the meaning of experiments in full. Most student
teachers praised the activities for providing an experience which was very close
to the actual process of design, performing and data analysis of a true research
experiment, but some complained about the excessive time spent on computer
based image processing and data analysis, with respect to time devoted to
hands-on activities. In this respect, further refinements of these experimental
activities, possible corrections on their positioning within the sequence, and
the collection of additional feedback from students are among our plans for
future work.
Q8: In the context of compact teacher training courses, is making the con-
ceptual and epistemological structure of quantum theory transparent and
easily verbalizable, through the Feynman approach, a viable strategy for
initiating appropriation, and providing teachers with the instruments and
motivation for proceeding in the study, and pursuing innovation in educa-
tional practice?
A very important result we obtained in both our tests with student teachers
was the positive evolution in the confidence they showed in talking and writing
about quantum physics. We do not here repeat again the nature of such evi-
dence, but in synthesis, most of their answers to explanation items in the final
tests demonstrated that they really were able to write about the subject, and in
particular about wave particle duality and the role of measurement, by draw-
ing information from their own mental models, and not by borrowing words
from external authorities. In the 2015 test, in particular, these satisfactory
result were obtained for about 80% of ST (most of those scoring a 3 or more
in Table 6.5). We consider this a very good result for a 10 hours course, with
a sample of students in which 63% had no previous knowledge about quantum
physics at all, and 26% had followed only one course in modern physics. This
is, clearly, not sufficient to assert that they will be able to teach quantum
physics, much the less that they will decide to teach it using an innovative or
research-based approach; but certainly can be considered a good first step in
providing them with the essential instruments, including a consistent and non
fragmented depiction of key concepts, to prosecute the study on their own.

Q9: In the high school context, can a teaching-learning sequence based on

Feynman’s formulation form the core nucleus of a complex, multi dimen-
sional learning environment meant to encourage appropriation of the sub-
ject matter by students? In this respect, is a comparison with other ap-
proaches to the teaching of quantum physics possible?

185
 7. Conclusions and future perspectives

As already stated in Section 6.3.5, our results concerning this research ques-
tion must, at the present time, be considered as very preliminary. Of the six
students which we interviewed, we analyzed the data of two, finding a clear
case of appropriation, and one in which, notwithstanding very good results
in tests, and also some significant degree of thickness of the student’s episte-
mological views on quantum physics, he did not accept it as a fundamental
theory, and maintained essentially an “hidden variable” view. Still, even in
this case, the Feynman model seemed to be extremely useful to the student,
as he provisionally accepted it as a source of knowledge and insight on the
theory, and did not resort to the construction of synthetic models. Although
we did not analyze the other four students’ interviews in detail, preliminary
data confirm that a learning environment which stimulates students on several
dimensions, including the historical-epistemological one, is really efficient in
leading students to intertwine the epistemology of quantum physics with their
own personal narrative; and that within this process, the Feynman approach
is of great value in helping students build consistent mental models.
An element of caution that must be mentioned, however, is that for two more
students which were not analyzed in detail, mental reserves similar to those of
Cheng seem to exist, not consisting in the production of alternative concep-
tions, but in the inner belief on the existence of a different, still undiscovered,
deterministic model. In one case such idea explicitly constructed by connect-
ing the students’ inner convictions to Einstein’s views, and appealing to his
intellectual authority.
While this confirms the idea that students have great difficulties in accepting
quantum theory, we do not want here to draw any further conclusions, not
even tentative ones, before comparing our data with those of other researchers
who have performed a similar work using an entirely different approach to the
disciplinary content, a work which will be carried our in the next months.

7.2 Future expansions of this work

A primary concern in extending the work contained in this dissertation should
be to reach the stage of a full validation [16, 17] of the teaching learning se-
quence. Although according to our data the proposal revealed its effectiveness
in classroom, the special context of the experimentation in which the teacher
was also designer is certainly a weakness of this study. Thus, the addressing
of transfer issues (dissemination of the proposal, formation of teachers to the
task of implementing it in their own classrooms, and new tests of the results)
must be considered as a first priority in order to maximize the impact of the
present work.
Although in designing our TLS we have chosen not to adopt an historical ap-
proach, the introduction, in the high school test, of an historical perspective
embodied in debates between great scientists of the past on the meaning of
the uncertainty principle and wave particle duality has certainly produced a

186
7.2. Future expansions of this work

relevant effect on students’ learning. On one hand, the discussion of different

historical perspectives has certainly helped students to disentangle, for exam-
ple, the currently accepted meaning of the uncertainty principle as intrinsic
indeterminacy from other possible interpretations. Furthermore, students’ ar-
gumentative papers reveal that the reflection on historical debates has helped
students in shaping their own personal view of quantum theory, and in develop-
ing a critical attitude towards science. On the other hand, as mentioned above
an analysis of the final interviews reveals that the authority exerted by Ein-
stein’s criticisms may have increased students’ skepticisms on quantum theory.
A more in depth and detailed evaluation and planning of the role, scope and
breadth that the introduction of elements from the history of physics should
have in the implementation of our TLS is certainly needed in view of novel
future tests.
At a lower priority level, our future plans also include expansions of the se-
quence itself. The amount of time necessary for the implementation of our TLS
is already near the limit of what is available in the final year of high school.
However, perhaps a very basic treatment of time evolution should be included.
In this perspective, two simple examples could be treated: the time evolution
of a wave function of the infinite well potential made of the superposition of
two stationary states, and the time evolution of a free wave packet. In both
cases, these problems would be discussed using simulations. These two simple
examples could serve, respectively, to deal with the possible wrong conception
of bound systems, one in which the stationary states are the only possible
ones [188]; and to show the functioning of Feynman’s path integral propagator
in a very simple case, similarly to what Taylor has done [5]. The connection
between the two examples can be built in terms of the information initially
available on the system, i.e. the initial state. In the first case, we know that
the quantum object can only be in one of two possible fixed energy states, so
we have only two energies to use for the evolution from the initial to the final
time, each one corresponding to one of the two possible initial states. In the
second case, the range of possible energies for the quantum object is continu-
ous, and so we, in principle, can use all of them to connect with possible paths
points (x, ti ) to which a non zero amplitude is associated in the initial wave
function, to points (x, tf ) of the final one.
The duration of our sequence is also at the limit of time available for the sub-
ject of quantum physics in teacher instruction in the current Italian system.
However, in the perspective of transforming the sequence in a true introduc-
tory “course” for non specialists, certainly a more detailed treatment of spin
than the sole explanation of the existence of bosons and fermions and the Pauli
exclusion principle is in order. In this case, an idea which we have been pon-
dering for a long time is to first build a connection between the two inputs
Mach-Zehnder and more general two state systems, similarly to what is done
in Refs. [293, 395]. Then, the version of the Mach-Zehnder with polarizers can
be introduced and, from there, the transition to the phenomenology of spin 12

187
 7. Conclusions and future perspectives

systems and the Stern-Gerlach experiment is relatively easy.

Finally, some of our plans for future research concern the experimental dimen-
sion. In particular, we are thinking of using the approach we developed to
spectroscopy with low cost instrument, to develop an experimental activity on
nuclear radioactivity and radiometric dating based on the analogy with the
atomic phenomenon of fluorescence.

188
Appendix A
Test sheets and questionnaires

In this Appendix we report the tests which we delivered to students in our three
experimentations. In the case of multiple choice items, the answer judged
as correct (if applicable) is reported in bold. For open questions, the space
students had available for the answer is specified.

A.1 2014 experimentation with student teach-

ers
A.1.1 Pre-test
1. During your studies, had you previously taken courses in modern physics?

A. No
B. Yes, one
C. Yes, more than one

2. In your experience as a secondary school teacher, have you ever taught

optics? If so, in what class and at what level?

A. No
B. Yes (please specify) (2 lines space)

3. How would you answer to a student of yours who asks “What is light?”.(8
lines space)

4. List what are, in your opinion, the main differences between classical physics
and quantum physics. (10 lines space)

5. What is, in your opinion, the meaning of Heisenberg’s uncertainty principle?

(8 lines space)

189
 A. Test sheets and questionnaires

6. Describe the interference phenomenon occurring

when monochromatic light passes through a double-
slit. It is not necessary to focus on the formulas, but
rather on the phenomenon as a whole. Is it possi-
ble to interpret the same phenomenon in terms of
photons? Explain your answer. (12 lines space)
7. What is in your opinion the meaning of “wave-particle duality”? (8 lines
space)
8. True or false: an electron runs approximately sinu-
soidal trajectories around the nucleus, such as those
shown in figure.

A. True

B. False

9. Why is not Heisenberg’s uncertainty principle normally taken into account

in the description of macroscopic bodies, such as marbles and tennis balls?
A. Because they are not disturbed by instruments revealing them, contrar-
ily to microscopic particles.
B. Because for macroscopic bodies the laws of quantum mechanics are not
valid.
C. Because macroscopic bodies are directly accessible to our senses, so the
uncertainty principle in this case is irrelevant.
D. Because, although the principle applies to macroscopic bodies
also, the resulting uncertainty is too small to be significant.

A.1.2 Post-test

1. In the idealized setup presented in figure, a pho-

ton is emitted by the source. A, B, C, D are
slits on screens, while two detectors are present
at points Z and W . The ratio P (Z)/P (W ) be-
tween the probabilities of detecting the photon
in Z and W is:

A. 1

B. 1/2
√
C. 2

D. 0

190
A.1. 2014 experimentation with student teachers

2. If, in the same setup as the previous item, the information is in some way
acquired, that the photon did not pass from slit C, the ratio P (Z)/P (W )
between the probabilities of detecting the photon in Z and W becomes:

A. 1
B. 1/2
√
C. 2
D. 0

3. In a double slit experiment using one electron at a time, which one of the
following statements is true?

A. It is possible to know through which slit each electron passes, while at

the same time maintaining the interference pattern, if the technology
used is sufficiently advanced.
B. It is possible to know through which slit each electron passes,
but in this way the interference pattern is inevitably lost.
C. Even if an interference pattern is observed on the screen, each electron
must have passed or from one of the slits or the other, though we are
unable to determine which one.
D. It is possible to know through which slit each electron passed, but in do-
ing so at least half of the electrons are disturbed in their path, reducing
the visibility of the interference pattern.

4. In a laboratory an electron and a neutron move at the same speed. What

can be said of their wavelengths?

A. λelectron > λneutron

B. λelectron < λneutron
C. λelectron = λneutron
D. Nothing, because wavelength is only a a meaningful quantity for pho-
tons.

5. A double slit experiment is performed using a very weak source, emitting

individual photons one at a time. First a normal experiment is performed.
Then, beyond one of the slits, a thin film of reflective material is placed,
as in the figure below. After accumulating a high number of photons, it
s observed that the fringes are inverted with respect to the case of the
normal experiment (minima in place of maxima, and vice versa). Provide
a qualitative interpretation of the phenomenon in terms of photons. Ignore
the effect of reflections on (and inside) the thin film. (12 lines space)

191
 A. Test sheets and questionnaires

6. List what are, in your opinion, the main differences between classical physics
and quantum physics. (10 lines space)
7. What is, in your opinion, the meaning of Heisenberg’s uncertainty principle?
(8 lines space)

A.1.3 Open question in the final exam

ˆ Acquiring information on a quantum system through a measurement usually
modifies its future evolution. Comment the previous statement, confronting
it with the classical case and referring to the two slit experiment, or to other
experimental situations which were discussed during the course. (Answer
length was unrestricted.)

A.2 2015 test with student teachers

A.2.1 Initial mini-questionnaire
1. During your studies, had you previously taken courses in modern physics?
A. No
B. Yes, one
C. Yes, more than one
2. List what are, in your opinion, the main differences between classical physics
and quantum physics. (10 lines space)

A.2.2 Final exam

1. In a double slit experiment using one electron at a time, which one of the
following statements is true?
A. It is possible to know through which slit each electron passes, while at
the same time maintaining the interference pattern, if the technology
used is sufficiently advanced.

192
A.2. 2015 test with student teachers

B. It is possible to know through which slit each electron passes,

but in this way the interference pattern is inevitably lost.

C. Even if an interference pattern is observed on the screen, each electron

must have passed or from one of the slits or the other, though we are
unable to determine which one.

D. It is possible to know through which slit each electron passed, but in do-
ing so at least half of the electrons are disturbed in their path, reducing
the visibility of the interference pattern.

2. An electron in an atom can occupy the energy

levels shown in the figure. The electron is
initially in the state with energy E3 . What
is the highest energy that an emitted photon
can have in this situation?

A. E4

B. E4 − E1

C. E3

D. E3 − E1

E. E3 − E2
3. In the Mach-Zehnder interferometer
represented in figure, the source emits
a single, monochromatic photon. The
two arms of the interferometer (BS1-
M 1-BS2 and BS1-M 2-BS2) have
the same length, as do the traits
BS2-Detector A and BS2-Detector
B. Consider detectors with 100% ef-
ficiency. In these conditions, it can
be assumed that a photon will be de-
tected:

A. Never, neither at detector A or B.

B. Certainly or almost certainly at A.

C. Certainly or almost certainly at B.

D. At A or B with 50% probability.

E. At both detectors a photon will be detected with half the original energy.

193
 A. Test sheets and questionnaires

4. In the Mach-Zehnder setup of the pre-

vious exercise, a third detector C is
added, as in the figure. Assume that
the detection in C is interaction-free,
so that the photon, even if revealed
at C, can continue on its path. Con-
sider detectors with 100% efficiency.
In these conditions, it can be assumed
that a photon will be detected:

A. Never, neither at detector A or B.

B. Certainly or almost certainly at B.
C. At both detectors a photon will be detected with half the original energy.
D. Certainly or almost certainly at A.
E. At A or B with 50% probability.

5. Acquiring information on a quantum system through a measurement usu-

ally modifies its future evolution. Comment the previous statement, con-
fronting it with the classical case and referring to the two slit experiment,
or to other experimental situations which were discussed during the course.
(Answer length was unrestricted.)

A.3 2015 test with high school students

A.3.1 Initial questionnaire
A.3.1.1 Questionnaire on the intuitive understanding of physical
theories - First part
Fill the following tables regarding laws and phenomena of classical physics
which you previously studied. Remember that the purpose of this activity is
not to assign you a mark, as some questions concern subjects that were treated
long ago. Feel free to write what you think, but still try to answer to the best
of your possibilities.

a. The Joule effect

1. Do you remember formulas or other elements of the mathematical model

corresponding to the named phenomenon? (6 lines space)
2. Can you draw images or visual schemes corresponding to the named
phenomenon? (Blank space corresponding to about 6 lines)
3. Can you describe in your words the phenomenon, its relevance, some of
its consequences or applications? (12 lines space)

194
A.3. 2015 test with high school students

4. How do you judge your current comprehension of the phenomenon?

A. Excellent
B. Good
C. Scarce
D. Null

b. The second principle of thermodynamics

1. Do you remember formulas or other elements of the mathematical model

corresponding to the named principle? (6 lines space)
2. Can you draw images or visual schemes corresponding to the named
principle? (Blank space corresponding to about 6 lines)
3. Can you describe in your words the principle, its relevance, some of its
consequences or applications? (12 lines space)
4. How do you judge your current comprehension of the principle?
A. Excellent
B. Good
C. Scarce
D. Null

c. Magnetic force on a current carrying wire

1. Do you remember formulas or other elements of the mathematical model

corresponding to the named phenomenon? (6 lines space)
2. Can you draw images or visual schemes corresponding to the named
phenomenon? (Blank space corresponding to about 6 lines)
3. Can you describe in your words the phenomenon, its relevance, some of
its consequences or applications? (12 lines space)
4. How do you judge your current comprehension of the phenomenon?
A. Excellent
B. Good
C. Scarce
D. Null

A.3.1.2 Questionnaire on the intuitive understanding of physical

theories - Second part
The following passages report reflections of physicists, or philosophers of sci-
ence, on the basic features of a theory. In particular, the themes discussed
are the meaning of the concept of “model” in physics; the relationship between
phenomena and their mathematical models; and the problem of whether such
models need or need not, in order to be comprehensible, admit an intuitive

195
 A. Test sheets and questionnaires

representation, in the form of images, of a narration, or of analogies. Read

carefully the passages, highlighting:
ˆ Words or sentences that are unclear
ˆ Keywords or expressions clarifying the views of the author.
1. H. Hertz: (from “Principles of mechanics presented in a new form”, 1899)

“The most direct, and in a sense the most important, problem which our
conscious knowledge of nature should enable us to solve is the anticipation
of future events, so that we may arrange our present affairs in accordance
with such anticipation.
As a basis for the solution of this problem we always make use of our knowl-
edge of events which have already occurred, obtained by chance observation
or by prearranged experiment. In endeavouring thus to draw inferences as
to the future from the past, we always adopt the following process. We form
for ourselves images or symbols of external objects; and the form which we
give them is such that the necessary consequents of the images in thought
are always the images of the necessary consequents in nature of the things
pictured.
In order that this requirement may be satisfied, there must be a certain
conformity between nature and our thought. Experience teaches us that
the requirement can be satisfied, and hence that such a conformity does in
fact exist. When from our accumulated previous experience we have once
succeeded in deducing images of the desired nature, we can then in a short
time develop by means of them, as by means of models, the consequences
which in the external world only arise in a comparatively long time, or as
the result of our own interposition. We are thus enabled to be in advance of
the facts, and to decide as to present affairs in accordance with the insight
so obtained.
The images which we here speak of are our conceptions of things. With the
things themselves they are in conformity in one important respect, namely,
in satisfying the above-mentioned, requirement. For our purpose it is not
necessary that they should be in conformity with the things in any other
respect whatever. As a matter of fact, we do not know, nor have we any
means of knowing, whether our conceptions of things are in conformity with
them in any other than this one fundamental respect.”

Questions:
a. What are some of the keywords of this passage? (2 lines space)
b. What does Hertz highlight in describing the fundamental aspects of
man’s knowledge about nature? (4 lines space)
2. J. C. Maxwell: (from “On Faraday’s lines of force”, 1861)

196
A.3. 2015 test with high school students

“The first process therefore in the effectual study of the science, must be
one of simplification and reduction of the results of previous investigation
to a form in which the mind can grasp them. The results of this simplifica-
tion may take the form of a purely mathematical formula or of a physical
hypothesis.
In the first case we entirely lose sight of the phenomena to be explained;
and though we may trace out the consequences of given laws we can never
obtain more extended views of the connexions of the subject. If, on the
other hand, we adopt a physical hypothesis we see the phenomena only
through a medium and are liable to that blindness to facts and rashness
in assumption which a partial explanation encourages. We must therefore
discover some method of investigation which allows the mind at every step
to lay hold of a clear physical conception without being committed to any
theory founded on the physical science from which that conception is bor-
rowed, so that it is neither drawn aside from the subject in the pursuit of
analytical subtleties nor carried beyond the truth by a favourite hypothe-
sis.
In order to obtain physical ideas without adopting a physical theory we
must make ourselves familiar with the existence of physical analogies. By
a physical analogy I mean that partial similarity between the laws of one
science and those of another which makes each of them illustrate the other.”

Questions:

a. What are some of the keywords of this passage? (2 lines space)

b. What does Maxwell highlight in describing the fundamental aspects of
man’s knowledge about nature? (4 lines space)

3. J. von Neumann: (from ”Method in the physical sciences”, 1961)

“To begin, we must emphasize a statement which I am sure you have heard
before, but which must be repeated again and again. It is that the sci-
ences do not try to explain, they hardly ever try to interpret, they mainly
make models. By a model is meant a mathematical construct which, with
the addition of some verbal interpretations describes observed phenomena.
The justification of such a mathematical construct is solely and precisely
that it is expected to work – that is correctly to describe phenomena from
a reasonably wide area.”

Questions:

a. What are some of the keywords of this passage? (2 lines space)

b. What does Von Neumann highlight in describing the fundamental as-
pects of man’s knowledge about nature? (4 lines space)

197
 A. Test sheets and questionnaires

4. H. W. De Regt and D. Dieks: (from ”A Contextual Approach to Sci-

entific Understanding”, 2005)

“Scientists are not unanimous about the nature of understanding either.

Lord Kelvin famously stated: “It seems to me that the test of “Do we or
not understand a particular subject in physics?’ is, “Can we make a me-
chanical model of it?””. But while this view of scientific understanding had
strong appeal and was widely supported in mid-nineteenth century, today
no physicist will defend it: classical mechanics has long lost its paradig-
matic position. As a second example, consider the fact that today hardly
any scientifically educated person will judge Newton’s law of inertia unin-
telligible, whereas to most of Newton’s contemporaries it was mysterious.
The history of science shows great variation in what is and what is not
deemed understandable. Even at one particular moment in history opin-
ions about what is understandable often diverge (...)
We propose a general Criterion for the Intelligibility of Theories:
A scientific theory is intelligible for scientists (in a given context) if they
can recognise qualitatively the characteristic consequences of the theory
without performing exact calculations. ”

Questions:

a. Do you agree with De Regt and Dieks’ proposal for defining a theory as
intelligible? Do you think their definition is usable in practice? (6 lines
space)
b. Do you see any connection between such proposal and the ideas of physi-
cists of the past which were reported in the previous passages? (8 lines
space)

In the following passage Douglas Hofstadter tells about his career as a student,
and his firs steps in research. Read the passage carefully, reflecting on what
you, in your experience as a student, find important for understanding a scien-
tific subject, referring in particular to physics and mathematics. At the end of
the reading, add some free comments about the thoughts that it has inspired
you.

5. D. R. Hofstadter (from “Analogies and metaphors to explain Gödel’s

theorem, 1982)”

“When I was a graduate student at Berkeley in mathematics during 1966

and 1967, I found out, to my chagrin, that mathematics was too abstract
for me. I had always thought I was a pretty abstract thinker, but what
I realize about that time in my life was that, in fact, all my thoughts are

198
A.3. 2015 test with high school students

very concrete. They all are based on images, analogies, and metaphors. I
really think only in concrete ideas, and I found that I couldn’t attach any
concrete ideas to some of the mathematics I was learning. I could learn
the formal statements and theorems, I could prove theorems formally, but
I really could not go beyond them. I was just not able to get the con-
cepts without images, so I turned away from mathematics and went on to
physics at the University of Oregon. Then my career went through var-
iegated phases, and finally I wound up in computer science and artificial
intelligence, which is not exactly an accident because my greatest inter-
est in artificial intelligence nowadays is in understanding analogies. In a
way I have come back to study, through computer science (and particularly
through the branch of it called artificial intelligence), what these analogies
are that I think with.”

Personal comments:(8 lines space)

Final questions:

a. As you will have noticed in reading the proposed passages, the authors
highlight the different roles that images, analogies, and mathematical
models play in the construction and understanding of physical theo-
ries. Referring to what you have previously studied in physics, can you
report specific examples illustrating their meaning, or demonstrating
other possible roles which these instruments of thought can play?(12
lines space)
b. Thinking about your personal experience in the study of physics, which
ones among these instruments (images, physical analogies and mathe-
matical models) have played an important role for your understanding
of the physical world?(10 lines space)

A.3.2 Guidelines for students’ argumentative paper

Composition of a “brief essay” or “newspaper article”.

Guidelines:1 Develop the subject reported below in the form of a “brief essay”
or “newspaper article” using all or a part of the accompanying documents.
If you choose the “brief essay” form, develop your argumentation making op-
portune reference to your personal knowledge and previous studies. Find a
coherent title for the essay and, if necessary, divide it in paragraphs.
If you choose the “newspaper article” form, specify the title and the kind of
newspaper on which you think it should be published.
For both types of composition, the maximum length should be five standard
1
The guidelines in this paragraph are standard for this type of task in the last year and
final exam of high school in Italy.

199
 A. Test sheets and questionnaires

half-page columns.

Subject: ‘Quantum physics and the representation of reality: possible path-

ways for imagination and intuition.”

1. N. Bohr, from “Discussion with Einstein on the epistemological problems

in atomic physics”, in Albert Einstein: Philosopher-Scientist, P. A. Schilpp,
ed., 1949.

Evidence obtained under different experimental conditions cannot be com-

prehended within a single picture, but must be regarded as complementary
(...). Under these circumstances an essential element of ambiguity is in-
volved in ascribing conventional physical attributes to atomic objects, as is
at once evident in the dilemma regarding the corpuscular and wave proper-
ties of electrons and photons, where we have to do with contrasting pictures,
each referring to an essential aspect of empirical evidence.

2. E. Schrödinger, quoted in Discussione sulla fisica moderna2 , Bollati Bor-

inghieri, 1959.

The picture of material reality is nowadays faltering and uncertain, as it

had not been for a long time. We know a crowd of very interesting details,
we learn of new ones weekly (...) . But the picture of the subject, which I
have to build before your eyes, does not exist at all, at the present moment,
there are only fragments, with a more or less partial truth value (...) Both
in the particle picture, and in the wave one, there is a content of truth,
which we can not give up. But we do not know how to merge these two
truths.

3. W. Heisenberg, from The Physical Principles of the Quantum Theory,

University of Chicago Press, 1930.

It is not surprising that our language should be incapable of describing

the processes occurring within the atoms, for, as has been remarked, it was
invented to describe the experiences of daily life, and these consist only of
processes involving exceedingly large numbers of atoms. Furthermore, it
is very difficult to modify our language so that it will be able to describe
these atomic processes, for words can only describe things of which we can
form mental pictures, and this ability, too, is a result of daily experience.
Fortunately, mathematics is not subject to this limitation, and it has been
2
The guidelines for students always contained the reference for the Italian version or
edition, while here we report the English edition. In this case, we could not find an English
reference for the quote, which was translated by us.

200
A.3. 2015 test with high school students

possible to invent a mathematical scheme — the quantum theory — which

seems entirely adequate for the treatment of atomic processes; for visual-
ization, however, we must content ourselves with two incomplete analogies:
the wave picture and the corpuscular picture.

4. A. Einstein and L. Infeld, from “The Evolution of Physics: From Early

Concepts to Relativity and Quanta.” Simon and Schuster, 1938.

Physical concepts are free creations of the human mind, and are not, how-
ever they may seem, uniquely determined by the external world. In our
endeavor to understand reality we are somewhat like a man trying to un-
derstand the mechanism of a closed watch. He sees the face and the moving
hands, even hears its ticking, but he has no way to open the case. If he
is ingenious he may form some picture of a mechanism which could be re-
sponsible for all of the things he observes, but he may never be quite sure
his picture is the only one which could explain his observations. He will
never be able to compare his picture with the real mechanism and he can-
not even imagine the possibility or the meaning of such a comparison. But
he certainly believes that, as his knowledge increases, his picture of reality
will become simpler and simpler and will explain a wider and wider range
of his sensuous impressions. He may also believe in the existence of the
ideal limit of knowledge and that it is approached by the human mind. He
may call this ideal limit the objective truth.

5. M. Weiss, from “Anschaulichkeit, Abscheulichkeit.” (Visualizability, re-

pulsiveness.) http://math.ucr.edu/home/baez/photon/anschaulichkeit.htm,
1992.

In the childhood of quantum theory, the matter of visualizability loomed

just as large. In his second paper on wave mechanics, Schroedinger wrote:
“...it has even been doubted whether what goes on in an atom can be
described within a scheme of space and time. From a philosophical stand-
point, I should consider a conclusive decision in this sense as equivalent to
a complete surrender. For we cannot really avoid our thinking in terms
of space and time, and what we cannot comprehend within it, we cannot
comprehend at all. There are such things but I do not believe that atomic
structure is one of them.“
Schroedinger wrote to Willy Wien:
“Bohr’s standpoint, that a space-time description is impossible, I reject a
limine... If [atomic research] cannot be fitted into space and time, then it
fails in its whole aim and one does not know what purpose it really serves.”
Bohr and Heisenberg of course held a different opinion. The founding pa-
pers on matrix mechanics expressed the operational philosophy: “You got
your equations, you got your observations, and they match. What more
do you want? Shut up and calculate!” Of course, they had to say it more

201
 A. Test sheets and questionnaires

politely, at least in print. For example, here’s the abstract, in full, of

Heisenberg’s famous paper:
“The present paper seeks to establish a basis for theoretical quantum me-
chanics founded exclusively upon relationships between quantities which in
principle are observable.”

6. J. M. Lévy-Leblond, from “On the nature of quantons.” Science and

Education, 2003.

That the true nature of quantum objects has long been misunderstood
is proved by their still all too common description in terms of an alleged
“wave-particle duality”. It must be remarked first of all that this formula-
tion is at best ambiguous. For it may be understood as meaning either that
a quantum object is at once a wave and a particle, or that it is sometimes
a wave and sometimes a particle. Neither one of these interpretations in
fact make sense. “Wave” and “particle” are not things but concepts, and
incompatible ones; as such, they definitely cannot characterise the same
entity. While it is true that quantum objects may in some cases look like
waves, and in other cases like particles, it is truer still that in most situa-
tions, particularly the ones explored by the elaborate modern experiments,
they resemble neither one nor the other. The situation here is reminiscent
of that encountered by the first explorers of Australia, when they discov-
ered strange animals dwelling in brooks. Viewed from the forefront, they
exhibited a duckbill and webbed feet, while, seen from behind, they showed
a furry body and tail. They were then dubbed “duckmoles”. It was later
discovered that this “duck-mole duality” was of limited validity, and that
the zoological specificity of these beasts deserved a proper naming, which
was chosen as “platypus”. Bunge’s proposal is to call them “quantons”.

7. P. A. M. Dirac, from “Principles of quantum mechanics.” Oxford Uni-

versity Press, 1930.

In answer to the first criticism it may be remarked that the main object
of physical science is not the provision of pictures, but is the formulation
of laws governing phenomena and the application of these laws to the dis-
covery of new phenomena. If a picture exists, so much the better; but
whether a picture exists or not is a matter of only secondary importance.
In the case of atomic phenomena no picture can be expected to exist in the
usual sense of the word ’picture’, by which is meant a model functioning
essentially on classical lines. One may, however, extend the meaning of the
word ’picture’ to include any way of looking at the fundamental laws which
makes their self-consistency obvious. With this extension, one may gradu-
ally acquire a picture of atomic phenomena by becoming familiar with the
laws of quantum theory.

202
A.3. 2015 test with high school students

A.3.3 Final test for school grading

1. A photon belonging to X-ray beam with energy E = 2.00 · 10−14 J is scat-
tered from a graphite target and is deviated by an angle α = 90. Calculate:

a) Its wavelength shift in the case it is scattered by an electron, which as

usual is hypothesized as initially at rest (me = 9.11 · 10−31 kg)
b) The kinetic energy acquired by the electron.

2. A source (an excited atom) emits a photon with a mean wavelength λ =

650 nm. Th intrinsic uncertainty on the instant of emission (i.e. the mean
lifetime of the excited state) is ∆t ' 10−9 s. Compute:

a) the uncertainty on the energy of the emitted photon.

b) The relative uncertainty on the wavelength of the emitted photon.

3. “Thermal” neutrons are neutrons with an average kinetic energy comparable

with the one of air molecules at room temperature,i.e. an energy E ' k · T ,
with T = 300 K. Calculate the De Brogle wavelength for thermal neutrons.
4. A Michelson interferometer is formed by two
mirrors and a beam splitter (with 50/50 re-
flection / transmission) set up as in the figure.

a) Identify the possible paths between the

Source (S) and the detector (D).

b) If the traits P -M 1 and P -M 2 are exactly

equal in length, calculate the expected
probability of finding a photon at D.

c) If the source emits green photons with

wavelength λ = 430nm find the minimum
variation in the distance P -M 2 needed to
obtain a maximum probability of detect-
ing a photon at D.
5. An electron beam with De Broglie wavelength λ = 2.0 · 10−7 m is directed
towards an unsurmountable obstacle with two slits, through which electrons
can pass, which are d = 1.0 mm apart. The electrons are then detected on
a screen which is D = 5.0 m distant from the slits.

a) Considering the slits as point-like, calculate the distance between two

successive maxima of the observed interference pattern, using the ap-
proximations of Fraunhofer and of small angles.
b) Considering now slits with very small but finite width (for example
a = 1.0 · 10−5 m) describe qualitatively what would be observed on
the screen if, on one of the slits, a detector would be placed, capable

203
 A. Test sheets and questionnaires

of revealing the passage of the electron. Precise calculations are not

necessary.

A.3.4 Final test-questionnaire

Item 1: Quantum uncertainty

In the textbook by U. Amaldi [421] (in the 2003 edition), the uncertainty
principle is presented in the following way:

In order to “see” a particle, we must have it scatter the light directed to-
wards it in such a way that part of the scattered light arrives to our eyes,
or to detection instruments. To do so, it is necessary that the wavelength
of light is at most equal to the length scale of the item which we desire to
“see”. (...)but the photons which form a light ray with small wavelength (...)

Figure A.1: Depiction of the uncertainty principle (reproduced from Ref.

[421]).

are very energetic and interact with material particles producing Compton
scattering.
In conclusion, the particle which we can see, because it has been hit by a
photon which has then arrived to our detector, has undergone a collision
which accelerated it in a random way. Thus, after the measurement, we
can know its position, but we lost all possibilities of precisely determining
its momentum. It is interesting to note that if we wish to have a lower
uncertainty ∆x on the position, using light with a lower wavelength, the
energy of incident photons must be increased, and as a consequence, the
uncertainty on the particle’s momentum increases.

Comment on this presentation, answering to the following questions:

204
A.3. 2015 test with high school students

a) What points are nowadays no longer acceptable as scientifically correct

in defining quantum uncertainty?
b) What criticisms can be raised to this introduction to uncertainty?
c) Which examples can be made to show that the discussion in the text-
book is not adequate?

Item 2: The Mach-Zehnder interferometer

In the Figures A.2 and A.3 the results of two possible experiments using
a Mach-Zehnder interferometer are shown. The first case is the ordinary
one, and the probability of detecting the photon is 100% at Detector B. In
the second case, using an intermediate detector C which detects the pas-
sage of the photon (without destroying it, and ideally without interacting
with it) the resulting probability is 50% for each one of the two detectors.

Figure A.2: The ordinary Mach-Zehnder experiment and its results (the two
arms of the apparatus have the same length).

After having described and analyzed (briefly, but also in a formal way) the
experiments in Figs. A.2 and A.3, answer the following questions:

a) Which properties of the quantum objects are highlighted in these ex-

periments?
b) How would you convince someone who does not know about quantum
physics that such aspects are surprising but not incomprehensible?

Item 3: Two slits experiment with one electron at a time

Referring to Figure A.4, discuss the following points:

205
 A. Test sheets and questionnaires

Figure A.3: Results for the Mach-Zehnder experiment with an intermediate de-
tector C (non interacting) which can reveal the passage of the photon without
destroying it.

Figure A.4: The Merli-Missiroli-Pozzi experiment 1974-76, reproduced from

Ref. [380]

a) If we modeled the electron as a wave, what would we expect to see on

the screen as the intensity of the beam is gradually lowered? Is such
model in contradiction with Fig. A.4, which was obtained experimen-
tally?

206
A.3. 2015 test with high school students

b) If we modeled the electron as a classical particle, the result seems to

be in contradiction with the principles of probability. Why? How can
the paradox be resolved through Feynman’s model? What is, for a
quantum object, the meaning of “summing amplitudes over paths?”
How is the probability of detecting a quantum object computed?
c) What conceptual instruments can be taken from the analysis of the
Mach-Zehnder experiment in Item 2 to interpret the Merli-Missiroli-
Pozzi experiment also?

207
 A. Test sheets and questionnaires

208
Appendix B
Exercises

In this Appendix we report the most significant exercises which were proposed
to both student teachers and high school students. In several cases, exercises
which were given as tests in one of the experimentations, were used as exercises
in another, so we only report problems which are significantly different from
the items previously presented in Appendix A.
Also, we only report exercises which we entirely constructed from scratch ex-
pressly for the sequence; in several cases, we proposed to students exercises
from textbooks or general repositories which we adapted or slightly modified
in the formulation in order to discuss them from the point of view of the sum
over paths approach; these exercises are not included in this Appendix.

B.1 Exercises for the 2014 version

1: Exercise on the sum over paths approach for the photon.

A photon is emitted by the source S with a definite wavelength λ. We

associate to the photon a vector ψ called amplitude which has initially
components (0, 1), i.e. a phase ψ = π2 (the choice of the initial phase is
arbitrary):

π
~
ψ(S) = (0, 1) = ei 2 = ↑ (B.1)

The slits A, B, P , Q are positioned as in Fig. B.1. Apart from the slits,
the screens are considered fully absorbing (black).

209
 B. Exercises

a) Identify the possible (piecewise rectilin-

ear, changing direction only at the slits)
paths between the Source (S) and each of
the slits (A, B, P , Q), and between the
source and detector (Z).

b) Using the sum over paths method, for

each of the points (A, B, P , Q, Z) com-
pute the length of each possible path,
draw the contribution (amplitude vector)
for each path, and the resulting amplitude
(sum of the vectors corresponding to each
path) i.e. the wave function at each point. Figure B.1: Setup for the first
part of the exercise. Point P is
symmetrical to Q with respect to
the vertical axis.

c) Assuming two detectors are placed in P

and Q as in Figure B.2, compute the ra-
tio between the probabilities of finding
the photon in P and Q.

d) Referring now again to Figure B.1, is the

probability of detecting a photon in Z
different from zero?

Figure B.2: Setup for question 1:c.

Again a photon is emitted by the source S with a definite wavelength λ
and the same phase choice as in Eq. B.1. The slits A, B, P , Q are now
positioned as in Fig. B.3. Apart from the slits, the screens are considered
fully absorbing (black).

210
B.1. Exercises for the 2014 version

e) Identify the possible (piecewise rectilin-

ear) paths between the Source (S) and
each of the slits (A, B, P , Q), and be-
tween the source and detector (Z).

f) Using the sum over paths method, for

each of the points (A, B, P , Q, Z) com-
pute the length of each possible path,
draw the contribution (amplitude vector)
for each path, and the resulting amplitude
(sum of the vectors corresponding to each
path) i.e. the wave function at each point.
Figure B.3: Setup for questions
1:e, 1:f and 1:h. Point P is sym-
metrical to Q with respect to the
z axis.

g) Assuming two detectors are placed in P

and Q as in Figure B.4, compute the ra-
tio between the probabilities of finding
the photon in P and Q. Compare the
result with the one of question 1:c.

h) Referring again to Figure B.3, is the

probability of detecting a photon in Z
different from zero? Compare the result
with the one of question 1:d.

Figure B.4: Setup for question 1:g.

i) In questions 1:c and 1:e the lengths of the vectors at P and Q are the
same. However, the final results for the detection probability at Z is
very different between the two cases (questions 1:d and 1:f). Thus, is
the length of the vector (without its phase) a sufficient information to
obtain a correct result for the detection probability?

211
 B. Exercises

j) Suppose now that an additional detector

is present only at slit P as in Fig. B.5, in
such a way that it can detect the passage
of the photon, before it arrives at Z. As-
sume that it does not detect the photon,
so that we know that it has passed from
Q. Draw, in this new case, all the pos-
sible paths from the source to Z. Is the
probability of detecting the photon in Z
different from zero? Repeat the calcula-
tion for the case that it is known that the
photon has passed from P . Compare the
two results with the case of question 1:h.
Figure B.5: Setup for question
1:j.

2: Exercise and activity on the Mach-Zehnder interferometer.

Consider the following experimental setup (Mach-Zehnder interferometer):

Figure B.6: Mach-Zehnder interferometer setup.

The arms of the interferometer are completely symmetric in length (i.e.

BS1-M 2 = M 1-BS2, BS1-M 1 = M 2-BS2, BS 00 -A = BS2-B).

a) Draw the possible paths of the photon from the source to detector A,
and from the source to detector B. Compute the probability of detecting
the photon at A and at B.
b) What is observed at detectors A and B in the case that either one of the
two mirrors M 1 and M 2 are removed from the setup? How can these
results be interpreted?

212
B.1. Exercises for the 2014 version

c) Returning to the full setup in Fig. B.6, assume that the source is a red
laser emitting photons with wavelength λ = 660 nm. Suppose that the
upper arm of the apparatus (BS1-M 1-BS2) is increased in length by
220 nm. Compute the probability that a photon reaches each one of the
two detectors.

Now explore the Geogebra simulation “machzehnder.ggb” of which a screen-

shot is reported in Fig B.7. Varying the parameter “d” means to increase
the width of a thin slit of material with refraction index n (which can also
be varied) interposed in one of the arms of the interferometer. In this case,
the simulation ignores the internal reflections in the film.

Figure B.7: Snapshot of the Mach-Zehnder simulation with the possi-

bility of including a dielectric film of variable length.

d) Observe the behaviour of the phasors corresponding to the amplitudes

at detector as the width of the film is varied. How can you interpret it
qualitatively, and formally?

3: Exercise on the normalization of probabilities.

The figure below (Fig. B.8) represents a two slits experiment in which,
rather than on a continuous screen, the quantum object (a photon in this
case) is revealed using seven discrete detectors P , Q, R, S, T , U , V .

213
 B. Exercises

Assume that the detectors are made

in such a way that the photon cannot
pass between two of them, but nec-
essarily be revealed by one of them.
However, we make the simplifying as-
sumption of using for computing am-
plitudes only the positions P , Q, R,
S, T , U , V which approximately cor-
respond to the centers of the de-
tectors. If λ is the wavelength of
the monochromatic photon emitted
by the source (which is equally distant
from the slits A and B), the lengths Figure B.8: Schematic representation of a
of the relevant segments in units of λ two slits experiment with discrete detec-
tors.
are:
AP = BV = 2 λ; BP = AV = 4 λ; AQ = BU = 2 λ; BQ = AU = 3.5 λ;
AR = BT = 2.25 λ; BR = AT = 3 λ; AS = BS = 2.5 λ.
Using these data, compute the detection probabilities at the detectors P ,
Q, R, S, T , U , V .

B.2 Exercises added in 2015

1: Exercise on three and four slits setups.

a) Consider a source emitting photons towards a screen on which three slits

are present, each one at a distance d from the successive one. Assume
the source can be considered equidistant from the three slits (in the
usual language of optics the source is “at infinity”), and that a screen is
placed at a large distance D from the slits. Find the maximum and min-
imum conditions for three slit interference. Are the maxima in the same
positions as in the analogous (i.e. with the same distance d between the
slits) system with only two slits? and the minima?
b) Consider now a grating with four slits, always with spacing d. The
maxima of detection probability are again at the same deflection angles?
And the minima? What do we find at those angles that correspond to
a minimum for the two slits system? And at those corresponding to a
minimum for the three slits system?

2: Exercise on probability and which way measurement using a three

slits system.

A photon with defined energy, and wavelength λ in the vacuum, is emitted

214
B.2. Exercises added in 2015

towards a system of three slits A, B, C. A screen is placed at a large dis-

tance from the slits (Fraunhöfer conditions) as represented in Fig. B.9. In
these conditions, the paths c1 , c2 , c3 can be considered as having the same
length, while paths c4 , c5 , c6 differ by λ/3 one from the successive (so that
c4 differs from c6 by 2λ/3).
a) Compute the ratio between the probabilities of
detecting the photon in Q and R.

b) Compute the same ratio in the case that at slit

B a non destructive detector is placed (which we
assume with 100% efficiency) that does not reveal
the passage of the photon at B.

c) Compute the same ratio in the case that the pas-

sage of the photon is revealed at slit B.

d) Compute the ratio between the frequencies with

which a photon is detected at R and at Q, if
a detector at slit B is present, independently of
whether it does or does not reveal the passage of Figure B.9: Three slits
the photon. Compare to the case of question 2:a setup.
and comment1 .

3: Exercise on probability and which way measurement using a sys-

tem of sucessive slits.

Consider the experimental setup described in Fig. B.10, in which a source

S emits a monochromatic photon of wavelength λ in the vacuum, two suc-
cessive screens with double slits A and B (first screen); C and D (second
screen) and a detector R after the second screen. Consider all the slits as
pointlike.

a) Draw all the possible (piecewise rectilinear, changing direction only at

slits) paths leading from S to R, and calculate the phase shifts between
the respective vectors. In this situation, the probabilty of detecting
the photon at R is (with respect to other points on the final screen)
maximum, minimum or neither maximum nor minimum?

1
Item 2:d was only included in this form in the experimentation with student teachers.
In the case of high school students only a qualitative comparison was required.

215
 B. Exercises

b) Consider now the case that at slit B

a detector is placed, which does not
reveal the passage of the photon. In
this new situation, what are the pos-
sible paths leadng from S to R? The
probabilty of detecting the photon at
R is (with respect to other points on
the final screen) maximum, minimum
or neither maximum nor minimum?

c) Finally, repeat the calculation in the

case that the detector in B reveals the
passage of the photon. Compare the
results of questions 3:b and 3:c with
those of question 3:a and comment. Figure B.10: Setup with multiple
slits for questions 3:a, 3:b and 3:c.

4: Introductory activity and exercise on the De Broglie wavelength.

Watch the video of the accumulation of electrons on the screen forming

a two slits interference pattern from the site ([385]). According to the
explanations reported on the site and the accompanying article, the inter-
ference pattern was obtained shooting electrons accelerated by a difference
of potential V = 600 V against two slits which were d = 273 nm apart one
from the other, and detected on a screen which was D = 240 mm distant
from the slits.

a) Using these data, compute the distance between two successive maxima
(and thus the length scale of the video you are observing) in Fraunhöfer
approximation.
b) The article also explains that the experiment was performed keeping
the entire apparatus at a pressure of about 10−5 P a, or 10−10 times the
normal atmospheric pressure. Can you explain qualitatively why this
measure was necessary?

5: Introductory activity and exercise on resonant quantum systems.

An electron with mass me and defined kinetic energy E is emitted from

a source along the horizontal axis. A detector is placed at a distance a + 2b
from the source, along the direction of flight of the electron. Two barri-
ers are placed in A and B at a distance a one from another, as in figure B.11.

216
B.2. Exercises added in 2015

Figure B.11: Setup for exercise 5.

The barrier behave as semi-reflecting mirrors, in the sense that each time
the path of the electron encounters a barrier, it can be transmitted or
reflected. Each transmission or reflection
√ involves the multiplication of the
vector associated to the path by 1/ 2. For each reflection, the phase of
the vector associated to the path is also shifted by −π.

a) In this setup, infinitely many paths exist leading from the source to the
detector. Explain why, and how these paths are formed.
b) Calculate, for a generic energy E of the electron, the phase difference
between two paths which differ by one back and forth trip between the
barriers.
c) now explore the simulation ”doublebarrier.ggb”. You will see that, by
varying the energy E of the incoming electron, the probability of detec-
tion at D oscillates strongly, and in particular a discrete set of values
of E exist, for which the probability has sharp peaks. Using the result
of question 5:b, and looking at the behavior of the vectors associated
to paths which is shown in the simulation, find an analytic condition
expressing the values of energy for which the probability of detection is
maximum.

217
 B. Exercises

218
Appendix C
Protocol for the final interviews
in the high school course

In this appendix we reproduce the protocol used for the final interviews to
selected students, discussed in Section 6.3.5 and its subsections. The structure
of the protocol is borrowed from the one used in Refs. [34, 396], and several
questions are the same. The total time available for interviews was 40-50
minutes. We only report the individual questions designed for the two, out
of six who were interviewed in total, students which were considered in the
preliminary analysis.

C.1 Final interview for the quantum physics

sequence
C.1.1 Introduction to the interview
In class you have followed an articulated learning path in quantum physics,
in which you progressively built, starting from the properties of the photon
and the processes of interaction between radiation and matter, the quantum
description of the photon itself, as well as of other objects such as the electron
and the atom. You have also discussed the need to move beyond the wave and
particle descriptions to construct a new concept of “quantum object”. In doing
this you have looked for new images and analogies to explore this concept and
describe the properties emerging from real and thought experiments that you
have studied. You also have introduced a minimal formalism and set of rules,
based on the sum on the paths, that allowed you to interpret the results of
some experiments. We would like this conversation to focus on some of these
concepts and phenomena (quantum object, uncertainty relation, the sum on
the paths, ...) and the way you gradually built these concepts and understood
their properties.

219
 C. Protocol for the final interviews in the high school course

C.1.2 List of questions to ask students:

A. Questions on specific concepts:

1. Thinking about the path, which are the concepts that you understood
better, and which worse? What difficulties did you have?
2. Again, thinking about what you learned in the path, that words would
you use to describe the quantum object? What in your opinion differ-
entiates it most from the classical object?
3. During the path you have encountered the sum of amplitudes over all the
paths in different experiments (for example, in the two slit experiment,
in the Mach-Zehnder interferometer, in Zhou-Wang-Mandel). Can you
tell us, for each case, what it means to sum of all possible paths? Are
there differences? How does the acquisition of information on a quantum
system enter in identifying the possible paths? How can this be used to
explain wave particle duality?
4. The uncertainty principle has been proposed in different ways: through
the mathematical relationship between position and momentum in sin-
gle slit diffraction; as the analogous relationship between time and en-
ergy, meant as the instant of emission of a photon and its energy; dis-
cussing the historical debate between Heisenberg and Bohr; and finally
as a principle explaining the fact that confined quantum objects always
have a minimum of energy. Which one among the above points of view
the most useful way for you to understand the meaning of quantum
indeterminacy, and its revolutionary character? Why? What made you
realize about it?
5. What is, in your opinion, the aspect, experiment or concept that more
than others helps you “enter” the world of quantum physics, showing
its “oddity” compared to classical physics? How would you convince
someone who does not know about quantum physics that this aspect is
surprising but understandable in the light of a new quantum logic?

B. Questions of a more general nature about the course:

1. Which ones, between images, analogies, models, experiments, formal

models, exercises are the tools that allowed you to better understand
the aspects of quantum physics?
2. Were the reflections stimulated by the initial questionnaire on physical
theories important to you, or you think they were unnecessary for intro-
ducing quantum physics, or for reconstructing the path in retrospect?
If they were important for you, what contribution did they give to your
reflection? If they have been unnecessary, could you suggest different
topics that could be highlighted instead, at the beginning or during the
course?

220
C.1. Final interview for the quantum physics sequence

3. Were the experimental activities which you performed in the laborato-

ries in Pavia important for your understanding? Why?
4. Do you think the course was stimulating for the class, and did it trigger
discussions among you? Why?
5. Did you notice any differences between the way the quantum physics
path has been proposed, compared to the way you dealt with other
subjects?
6. Did you have different expectations about the path, for how it was
presented at the beginning?

C. Specific individual questions

ˆ To Chiara:
1. In your composition tou described non-epistemic uncertainty in quan-
tum physics in particularly effective terms: “Uncertainty is inherent to
the objects of the quantum world, and is that only when a physical mea-
surement is performed that it is possible to acquire a real value, but as
long as the measurement is not made, the quantum object remains in a
state that is objectively indefinite, and describes only the “potentiality”
of the object.” Did this issue particularly interest you, did you feel the
need to delve deeper into it?
2. In recent times you have expressed the concern that science from the
1900 is “abandoning logic”. Would you like to tell us more about this
preoccupation?
3. In the final test, after analyzing the Mach-Zehnder setup in the two cases
of presence and absence of a which way detector detector, you wrote that
the results obtained are surprising from a classical point of view, but
not incomprehensible, because “they can be proved mathematically and
visibly considering certain characteristics as inherently true.” In the
end, did the quantum model satisfy your need of mathematical rigor
and logic, or didn’t? When you write “visibly and mathematically”,
can it be understood that, according to you, for a phenomenon to be
understandable, it must, in addition to a being expressed through a
consistent mathematical model, permit some form of visualization?
ˆ To Cheng:
1. In the composition you wrote, among several other subjects, about
entanglement, writing: “two particles remain connected as they move
apart, continuing to send each other instantaneous information, faster
than light, a fact that is impossible in the theory of relativity.” You also
wrote about the Einstein’s doubts about this phenomenon. This is a
topic that was not treated in the sequence, so is it a personal research
that you made in relation to the quantum physics path, or independently
of it? Why did this phenomenon interest you?

221
 C. Protocol for the final interviews in the high school course

2. In the composition and the final test, when discussing the idea that the
acquisition of which way information reduces the possible paths, in both
cases you discussed the Zhou-Wang-Mandel experiment. Could you now
exemplify the concept referring to a much simpler example, namely the
double slit experiment with or without an ideal, non-interacting detector
placed at one of the slits?
3. In the final test, after analyzing the Mach-Zehnder experiment in the
two cases presented, you wrote that “the results may seem surprising,
but only because they are calculated differently” (in a case using the
quantum probability rules, in the other, the distinguishable case, the
classical rules); and that they are not incomprehensible, however, be-
cause in both cases a rule exists that allows you to calculate the results.
Can it be inferred that for you, the most important factor that makes
a phenomenon is understandable is the existence of consistent mathe-
matical model that describing it?

222
Appendix D
Realization and calibration of a
low cost spectrophotometer

We describe the construction and calibration of two slightly different kinds of

spectrometers, the first one using a transmission diffraction grating with 570
lines per millimetre (Fig. D.1 Left), the other one with a reflection diffraction
grating, made with a Compact Disk (Fig. D.1 Right). The apparatuses are
assembled by means of a black cardboard frame. A collimating slit produces
a narrow beam of parallel light. The gratings disperse the beam of light into
spectral lines, scattering different colors at different angles. The images of the
spectra are acquired by means of a photo camera, then analyzed using the
Tracker [56] video analysis software.
The calibration procedure for the spectrophotometers requires two steps:
wavelength calibration (λ-calibration) and intensity calibration (I-calibration).
The two steps are tightly connected, because only the calibration of the mea-
sured light intensity as a function of wavelength allows an accurate determina-
tion of the positions of the peaks of spectral lines. In fact, line widths may be
broadened due to temperature and pressure effects, but their peak is usually
still clearly visible if an intensity measurement is available.

D.1 Wavelength calibration

We use the light provided by a commercial fluorescent lamp, i.e. a low pressure
mercury-vapour gas-discharge lamp that uses fluorescence to produce visible
light. In these devices the fluorescent tube contains an atmosphere of an in-
ert gas, usually argon, krypton, neon, and a small amount of mercury vapour.
The discharge conditions in commercial lamps are quite standardized; gas pres-
sure in the tube is very low, about 3 Torr, and gas temperature stays within
320 − 400 K [424]. In these conditions, both pressure and Doppler broaden-
ings of spectral lines are completely negligible compared to the accuracy of
the apparatus. The spectrum shows many peaks distributed over the spec-

223
 D. Realization and calibration of a low cost spectrophotometer

(a) (b)

(c) (d)

Figure D.1: (a) and (b) The spectrometer set up: parallelepiped cardboard
with a hole for the lens of the camera and a narrow slit made affecting the
cardboard with a cutter. A lamp is placed in front and properly centered to
the collimating slit. The first order symmetrical spectra obtained with the
transmission diffraction grating are shown in (c), while the second and third
order spectra from the reflecting grating are shown in (d).

trum of visible light, due to halophosphate coating that contains europium

and terbium ions; and depending on the lamp type more or less power may
be contained in the continuous part of the spectrum. However, the principal
lines of the mercury spectrum at 436 nm (blue) and 536 nm (green) are al-
ways visible. We acquire pictures of the spectral lines obtained with the two
spectrometers and accurately measure their positions by using Tracker data
(which are based on pixel brightness) to determine the light peaks.
The value of λ corresponding to each peak is obtained by comparison with the
values measured with a high resolution commercial spectrometer [425]. For
the calibration of both spectrometers we used two lines, the blue and green
ones whose presence does not depend on the particular coating used in the
lamp. More in detail, calibration procedure consists in measuring the distance
in pixel between the chosen blue and green peaks, compute the proportional-
ity constant between wavelength and pixel number which we then use for all
spectrum lines. This corresponds to use a linear λ scale in the range 400 to
700 nm. We calculated that the maximum theoretical error induced by this

224
D.2. Intensity calibration

Table D.1: Wavelengths for fluorescent lamp obtained with the grating and
the CD spectrometers. In the second column the values obtained with an
Ocean Optics HR2000 spectrometer for the same kind of fluorescent lamp are
reported [5]. With (*) we indicate the peaks used for calibration.
Wavelength Species producing peak Grating spec- CD spectrome-
Colour of peak trometer λ of ter
(nm) peak (nm) λ of peak (nm)
Blue 436.6 mercury 436.5* 436.5*
Cyan 487.7 terbium from Tb3+ 488±2 488±2
Green 542.4 terbium from Tb3+ 543±2
546.5 mercury 544.5∗ 546.5*
Yellow 577.7 likely terbium from 577±2
Tb3+ or mercury
580.2 mercury or terbium from 579±2
Tb3+
584.0 possibly terbium from 581±2
Tb3+ or europium in
Eu+3 :Y2 O3
587.6 likely europium in 587±10 585±2
Eu+3 :Y2 O3
593.4 likely europium in 590±2
Eu+3 :Y2 O3
599.7 likely europium in 600±2
Eu+3 :Y2 O3
Red 611.6 europium in Eu+3 :Y2 O3 611±10 607±3

625.7 likely terbium from Tb3+ 625±3

approximation is less than 2 nm for λ in the range 400 − 600 nm and only for
the red lines becomes 5 − 7 nm for the transmission grating; the error is less
than 1 nm for the reflection grating on the whole 400 − 700 nm range. Results
obtained are reported in Table D.1, where an asterisk labels the peaks used
for calibration while the other peaks are used to confirm the reliability of the
measurement. The spectrometer using the CD reflection grating has a better
resolution and allows to distinguish the two green lines which differ by 2.9 nm
(see Fig. D.2 (a)).

D.2 Intensity calibration

Intensity calibration is necessary to measure the real light intensity which is
obtained starting from the RGB color space. The color system constructs all
the colors from the combination of Red, Green and Blue. The Red, Green and
Blue use 8-bits each, which have integer values from 0 to 255. This procedure
allows using our simple apparatus as a spectrophotometer. To obtain the
measure of light intensity from RGB values for a typical gamma-corrected
image generated by a photo camera, the procedure is well known [426]: first

225
 D. Realization and calibration of a low cost spectrophotometer

divide R, G and B by 255, and compute

Ymeasured = 0.2126 Rγ + 0.7152 Gγ +0 .0722 Bγ (D.1)

The γ = 2.2 exponent is needed to invert the gamma-correction operation

made by the camera, i.e. represents the relationship between pixel’s numerical
value and its actual luminance; and Y is a measure of luminance, which is
proportional to emitted light intensity per unit area. However the value ob-
tained depends on the response of the apparatus (CCD camera, lenses and so
on). So we have to evaluate the relative sensitivity [427] (spectral response)
of the apparatus by measuring the light spectrum from a known source. We
use known data for the spectral irradiance, ISun (λ) of the solar disk at Earth’s
surface [428] and compare them with our measurements of the solar spectrum,
YSun (λ) (averaged on many trials). Thus we experimentally obtain the relative
sensitivity,
ISun (λ)
R(λ) = (D.2)
YSun (λ)
which allows to convert the measured luminance in the actual intensity, in
arbitrary units, for any source. The function R(λ) is shown in Fig. D.2 (b). If
desired, one could then use I(λ) to measure with higher accuracy the positions
of the spectral lines, and improve the accuracy the λ-calibration. However the
refined measurements of the peak positions based on intensity differ from the
raw ones, based on brightness, by less than 2 nm.

(a) (b)

(c)

Figure D.2: (a) Spectra of a fluorescent lamp measured with the transmission
grating spectrophotometer (left) and the reflection one (right) and corrected
using R(λ) are compared with the analogous plot obtained with a professional
spectrophotometer (reported in the inset). (b) Relative sensitivity function
R(λ). (c) The spectrum of the fluorescent lamp used for calibration.

226
D.3. Comments

D.3 Comments
After completing the calibration procedure, the low cost spectrophotometers
we are described are capable of performing wavelength and intensity measure-
ments with good accuracy. In our educational path, they allow students and
teachers to observe discrete spectra from gas discharge lamps and, if an hydro-
gen gas lamp is available, to measure the lines of the Balmer series and evaluate
the Rydberg constant to less than 1% difference from its accepted value [59].
They can also be used in the experimental activity aimed at the measurement
of h described in Section 5.3.2 to measure the wavelength of light emitted by
LEDs, if the specifics are not available; and to study other phenomena related
to atomic emission spectra, such as fluorescence [429].

227
D. Realization and calibration of a low cost spectrophotometer

228
Bibliography

[1] De Ambrosis, A. and Levrini, O. (2010). “How physics teachers approach

innovation: An empirical study for reconstructing the appropriation path
in the case of special relativity.” Physical Review Special Topics-Physics
Education Research, 6(2), 020107.

[2] Feynman, R. P. (1948) “Space-time approach to non-relativistic quantum

mechanics.” Reviews of Modern Physics, 20(2), 367.

[3] Feynman, R. P. and Hibbs, A. R. (1965). Quantum mechanics and path

integrals. New York, NY: MacGraw Hill.

[4] Feynman, R. P. (1985). QED: The strange theory of light and matter
Princeton, NJ: Princeton University Press.

[5] Taylor, E. F., Vokos, S., O’Meara, J. M. and Thornber, N. S. (1998).

“Teaching Feynman’s sum-over-paths quantum theory.” Computers in
Physics, 12, 190.

[6] Dobson, K. , Lawrence, I. and Britton, P. (2000). “The A to B of quantum

physics.” Physics Education, 35(6), 400.

[7] Ogborn, J. and Taylor, E. F. (2005). “Quantum physics explains New-

ton’s laws of motion.” Physics Education, 40(1), 26.

[8] Hanc, J. and Tuleja, S. (2005). “The Feynman Quantum Mechanics with
the help of Java applets and physlets in Slovakia.” In 10th Workshop on
multimedia in physics teaching and learning, October 2005.

[9] Cuppari, A., Rinaudo, G., Robutti, O. and Violino, P. (1997). “Gradual
introduction of some aspects of quantum mechanics in a high school
curriculum.” Physics Education, 32(5), 302–308.

[10] de los Ángeles Fanaro M., Otero M. R., and Arlego M. ,(2012).“Teach-
ing Basic Quantum Mechanics in Secondary School Using Concepts of
Feynman Path Integrals Method.” The Physics Teacher, 50(3), 156–158.

229
 Bibliography

[11] Ogborn, J. and Whitehouse, M. (eds), (2000). Advancing Physics AS.

Bristol (UK): Institute of Physics Publishing.
[12] Styer, D. F. (2000). The strange world of quantum mechanics. New York
(NY): Cambridge University Press.
[13] Ogborn, J. (2006). “A First Introduction to Quantum Behav-
ior” http://www.if.ufrj.br/˜pef/aulas seminarios/notas de aula/carlos
2011 1/ensinoMQ/Ogborn QuantumBehavior.pdf.
[14] Hanc, J. (2006). “The time-independent Schrödinger equation in the
frame of Feynman’s version of quantum mechanics.” In Proceedings of
11th Workshop on Multimedia in Physics Teaching and Learning, Uni-
versity of Szeged, Hungary.
[15] Stefanel, A. (2008). “Impostazioni e percorsi per l’insegnamento della
meccanica quantistica nella scuola secondaria,” Giornale di fisica, 49(1),
15-53.
[16] Méheut, M., and Psillos, D. (2004). “Teaching–learning sequences: aims
and tools for science education research.” International Journal of Science
Education, 26(5), 515-535.
[17] Méheut, M. (2004). “Designing and validating two teaching–learning se-
quences about particle models.” International Journal of Science Educa-
tion, 26(5), 605-618.
[18] Garrison, J. C. and Chiao, R. Y. (2008). Quantum optics. Oxford (UK):
Oxford University Press.
[19] Onorato, P. (2011). “Low-dimensional nanostructures and a semiclassi-
cal approach for teaching Feynman’s sum-over-paths quantum theory.”
European Journal of Physics, 32(2), 259.
[20] Ireson, G. (1999). “A multivariate analysis of undergraduate physics
students’ conceptions of quantum phenomena.” European Journal of
Physics, 20(3), 193.
[21] Seifert, S., and Fischler, H. (1999). “A multidimensional approach for an-
alyzing and constructing teaching and learning processes about particle
models.” in M. Komorek, Behrendt, H., Dahncke, H., Duit, R., Graeber,
W., Kross, A. Research in Science Education-Past, Present, and Future,
2, 393-395.
[22] Steinberg, R., Wittmann, M. C., Bao, L. and Redish, E. F. (1999).“The
influence of student understanding of classical physics when learning
quantum mechanics.” In Papers presented at the annual meeting Na-
tional Association for Research in Science Teaching. March, 1999 (p.
41).

230
[23] Fischler, H. and Lichtfeldt, M. (1992). “Modern physics and students’
conceptions.”. International Journal of Science Education, 14(2), 181–
190.

[24] Greca, I. M., and Freire Jr, O. (2003). “Does an Emphasis on the Con-
cept of Quantum States Enhance Students’ Understanding of Quantum
Mechanics?” Science and Education, 12(5-6), 541-557.

[25] Tsaparlis, G. and Papaphotis, G. (2009). “High school Students’ Concep-

tual Difficulties and Attempts at Conceptual Change: The case of basic
quantum chemical concepts.” International Journal of Science Education,
31(7), 895–930.

[26] Vosniadou, S. (2012) “Reframing the classical approach to concep-

tual change: Preconceptions, misconceptions and synthetic models,” In
Second international handbook of science education. Dordrecht (NL):
Springer Netherlands, pp. 119-130.

[27] Vosniadou, S., Vamvakoussi, X. and Skopeliti I. (2008). “The framework

theory approach to the problem of conceptual change.” in Vosniadou, S.
(ed) International handbook of research on conceptual change. London
(UK): Routledge, pp. 3-34.

[28] Vosniadou, S., Ioannides, C., Dimitrakopoulou, A., and Papademetriou,

E. (2001). “Designing learning environments to promote conceptual
change in science.” Learning and instruction, 11(4), 381-419.

[29] Schrödinger, E. (1928). Collected Papers on Wave Mechanics. London:

Blackie, pp. 26-31.

[30] Heisenberg, Werner. (1927). “Über den anschaulichen Inhalt der quan-
tentheoretischen Kinematik und Mechanik.” Zeitschrift fr̈ Physik 43:
172–98.

[31] de Regt, H. W. (2014) “Visualization as a Tool for Understanding,”,

Perspectives on science, 22(3), 377-396.

[32] Levrini, O., Fantini, P. and Pecori, B. (2008). “The problem is not under-
standing the theory, but accepting it: a study on students’ difficulties in
coping with Quantum Physics,” In R. Jurdana-Sepic R., V. Labinac, M.
Zuvic-Butorac, A. Susac (Eds.), GIREP-EPEC Conference, Frontiers of
Physics Education (2007, Opatija), Selected contributions, pp. 319–324.

[33] Abhang, R. Y. (2005). “Making introductory quantum physics under-

standable and interesting.” Classroom-Resonance Journal of Science Ed-
ucation, 10, 63-73.

231
 Bibliography

[34] Levrini, O., Fantini, P., Tasquier, G., Pecori, B., and Levin, M. (2015).
“Defining and operationalizing appropriation for science learning,” Jour-
nal of the Learning Sciences, 24(1), 93-136.

[35] Levrini, O., and Fantini, P. (2013). “Encountering productive forms of

complexity in learning modern physics,” Science and Education, 22(8),
1895-1910.

[36] Borghi, L. and De Ambrosis, A. (2010). “Ricerca in Didattica della Fisica:

temi e problemi.” in Storia, Didattica, Scienze: Pavia 1975-2010. Pavia
(IT): Edizioni dell’Università degli Studi di Pavia.

[37] Psillos, D., and Méheut, M. (2001). “Teaching-learning sequences as a

means for linking research to development.” In Proceedings of the third
international conference on science education research in the knowledge
based society. Vol. 1, p. 226.

[38] Kattmann, U., Duit, R., Gropengießer, H., and Komorek, M. (1996).
“Educational reconstruction. Bringing together issues of scientific clari-
fication and students’ conceptions.” In Annual Meeting of the National
Association of Research in Science Teaching (NARST), St. Louis (1996,
April).

[39] Komorek, M., and Duit, R. (2004). “The teaching experiment as a pow-
erful method to develop and evaluate teaching and learning sequences
in the domain of non linear systems.” International Journal of Science
Education, 26(5), 619-633.

[40] Besson, U., Borghi, L., De Ambrosis, A., and Mascheretti, P. (2010).
“A Three-Dimensional Approach and Open Source Structure for the De-
sign and Experimentation of Teaching-Learning Sequences: The case of
friction.” International Journal of Science Education, 32(10), 1289-1313.

[41] Besson, U., Borghi, L., De Ambrosis, A., and Mascheretti, P. (2010).”
A Model of teacher preparation aimed at favouring the diffusion of re-
search based teaching practice.” Contemporary Science Education Re-
search, 101.

[42] Borghi, L., De Ambrosis, A., Lamberti, N., and Mascheretti, P. (2005).
“A teaching - learning sequence on free fall motion.” Physics education,
40(3), 266.

[43] Besson, U., De Ambrosis, A., and Mascheretti, P. (2010). “Studying the
physical basis of global warming: thermal effects of the interaction be-
tween radiation and matter and greenhouse effect.” European journal of
Physics, 31(2), 375.

232
[44] De Ambrosis, A., Malgieri, M., Mascheretti, P., and Onorato, P. (2015).
“Investigating the role of sliding friction in rolling motion: a teaching
sequence based on experiments and simulations.” European Journal of
Physics, 36(3), 035020.

[45] Onorato, P., and De Ambrosis, A. (2013). “How can magnetic forces
do work? Investigating the problem with students.” Physics Education,
48(6), 766.

[46] Onorato, P., Malgieri, M. and De Ambrosis, A. (2014). “Learning about

magnetic force: experiments versus theoretical explanations.” ESERA
2013 Conference, 2-7 September 2013, Nicosia, Cyprus.

[47] De Ambrosis, A., and Levrini, O. (2007). “Insegnare relatività ristretta

a scuola: esigenze degli insegnanti e proposte innovative.” Giornale di
Fı́sica, 48(4), 255-276.

[48] http://www.fisica.uniud.it/pls3.htm

[49] Wheeler, J. A., and Taylor, E. F. (1966). Spacetime physics. New York
(NY): W.H. Freeman.

[50] Resnick R. (1968). Introduction to Special Relativity. New York (NY):

John Wiley and Sons.

[51] Borghi L., De Ambrosis A, Mascheretti P., Massara C.I. (1987). “Com-
puter Simulation and laboratory work in the teaching of mechanics” -
Physics Education, 22, 177.

[52] Borghi L., De Ambrosis A., Gazzaniga G., Ironi L., Mascheretti P., Mas-
sara C.I. (1991). “Practical use of simulations to study relative motion.”
Computers and Education, 16(2), 157.

[53] Onorato, P., Malgieri, M. and De Ambrosis, A. (2015). “Rolling motion:

experiments and simulations focusing on sliding friction forces.” Proceed-
ings of the GIREP-MPTL Conference, Palermo, 7-12 July 2014.

[54] Onorato P, Mascheretti P. and De Ambrosis A. (2010). “Mechanical sen-

sors and plastic syringes to verify the gas laws without neglecting fric-
tion.” Physics Education 45, 586-593.

[55] Onorato P., Mascoli D. and De Ambrosis A. (2010). “Damped oscillations

and equilibrium in a mass-spring system subject to sliding friction forces:
Integrating experimental and theoretical analyses.” American Journal of
Physics 78, 1120-1127.

[56] https://www.cabrillo.edu/˜dbrown/tracker/

233
 Bibliography

[57] Onorato, P., Mascheretti, P., and DeAmbrosis, A. (2012). “Investigat-

ing the magnetic interaction with Geomag and Tracker Video Analy-
sis: static equilibrium and anharmonic dynamics.” European Journal of
Physics, 33(2), 385.
[58] Malgieri, M., Onorato, P., Mascheretti, P., and De Ambrosis, A. (2013).
“Reconstruction of Huygens’ gedanken experiment and measurements
based on video analysis tools.” European Journal of Physics, 34(5), 1145.
[59] Onorato, P., Malgieri, M. and De Ambrosis, A. (2015) “Measuring the
hydrogen Balmer series and Rydberg’s constant with a homemade spec-
trophotometer.” European Journal of Physics 36, 058001.
[60] Bevilacqua F., Bonera G., Borghi L., De Ambrosis A. and Massara C.I.
(1990). “Computer simulations and historical experiments.” European
Journal of Physics, 11, 15-24.
[61] Galili, I. (2008). “History of Physics as a tool for teaching,” in Vicentini,
M. and Sassi, E., (Eds.) Connecting Research in Physics Education with
Teacher Education. International Commission on Physics Education.
[62] Besson U and De Ambrosis A. (2011) “Teaching energy concepts by
working on themes of cultural and environmental value.” In Seroglou
F., Koulountzos V. and Siatras A. (eds.) Proceedings of the 11th IHPST
Conference. Athens (GR): Epikentro Publications.
[63] Besson, U., and De Ambrosis, A. (2014). “Teaching energy concepts by
working on themes of cultural and environmental value.” Science and
Education, 23(6), 1309-1338.
[64] Onorato, P., and De Ambrosis, A. (2012). “Particle tracks in a cloud
chamber: historical photographs as a context for studying magnetic
force.” European Journal of Physics, 33(6), 1721.
[65] Onorato, P., Mascheretti, P., and De Ambrosis, A. (2013). “Studying mo-
tion along cycloidal paths by means of digital video analysis.” European
Journal of Physics, 34(4), 921.
[66] Malgieri, M., Onorato, P., Mascheretti, P., and De Ambrosis, A. (2014).
“Pre-service teachers’ approaches to a historical problem in mechanics.”
Physics Education, 49(5), 500.
[67] Besson, U. (2010). “Calculating and understanding: Formal models and
causal explanations in science, common reasoning and physics teaching.”
Science and Education, 19(3), 225-257.
[68] Giuliani, G. (2001). “Heinrich Hertz: fisica, metodo e filosofia.” in
Guidone M., Muzzarelli Formentini C.,(eds), Roberto Clemens Galletti
di Cadillac pioniere della telegrafia senza fili Fermo (IT) (pp. 111-119).

234
[69] Giuliani, G. (2010). “Vector potential, electromagnetic induction and
“physical meaning”.” European Journal of Physics, 31(4), 871.
[70] Balzano, E., Guidoni, P., Minichini, C. (2007) “I modi del pensare:
comune, fenomenologico, per modelli.” In: Approcci e proposte per
l’insegnamento-apprendimento della fisica a livello pre-universitatio dal
Progetto PRIN-F21. Udine (IT): Forum (pp. 61-65).
[71] Balzano, E., Gagliardi, M., Giordano, E., Guidoni, P., Minichini, C., and
Tarsitani, C. (2008). “Lo studio delle onde dalla scuola dell’infanzia al
termine della scuola secondaria superiore.” Forum.
[72] Engle, R.A. and Conant, F. R. (2002). “Guiding Principles for Foster-
ing Productive Disciplinary Engagement: Explaining an Emergent Argu-
ment in a Community of Learners Classroom.”Cognition and Instruction,
20:4, 399-483.
[73] Cobb, P., Gresalfi, M., and Hodge, L. L. (2009). “An interpretive scheme
for analyzing the identities that students develop in mathematics class-
rooms.” Journal for Research in Mathematics Education, 40(1), 40-68.
[74] Nasir, N.S. and Hand, V. (2008). “From the Court to the Classroom:
Opportunities for Engagement, Learning, and Identity in Basketball and
Classroom Mathematics.” Journal of the Learning Sciences, 17:2, 143-
179.
[75] Levrini, O. (2014) “How can the Learning of Physics support the
Construction of Students’ Personal Identities?.” Talk presented at the
GIREP-MPTL Conference, Palermo, July 10th , 2014.
[76] Kalkanis, G., Hadzidaki, P. and Stavrou, D. (2003). “An instructional
model for a radical conceptual change towards quantum mechanics con-
cepts,” Science Education, 87(2), 257–280.
[77] Ayene, M., Kriek, J. and Damtie, B. (2011).“Wave-particle duality and
uncertainty principle: Phenomenographic categories of description of
tertiary physics students’ depictions.” Physical Review Special Topics-
Physics Education Research, 7(2), 020113.
[78] Fletcher, P., and Johnston, I. (1999). “Quantum mechanics: exploring
conceptual change.” In Papers presented at the annual meeting National
Association for Research in Science Teaching March, 1999 (p. 28).
[79] Kuhn, T. S. (1962) The structure of scientific revolutions. Chicago (IL):
University of Chicago press, 2012.
[80] McCloskey, M., Caramazza, A., and Green, B. (1980). “Curvilinear mo-
tion in the absence of external forces: Naive beliefs about the motion of
objects.” Science, 210(4474), 1139-1141.

235
 Bibliography

[81] Posner, G. J., Strike, K. A., Hewson, P. W., and Gertzog, W. A. (1982).
“Accommodation of a scientific conception: Toward a theory of concep-
tual change.” Science education, 66(2), 211-227.

[82] Posner, G. J., and Gertzog, W. A. (1982). “The clinical interview and the
measurement of conceptual change.” Science Education, 66(2), 195-209.

[83] Vosniadou, S. (2014). “The interplay of domain-specific and domain gen-

eral processes, skills and abilities in the development of science knowl-
edge.” Talk presented at the 2014 1st International Conference on: New
Developments in Science and Technology Education May, 29-31, Corfu,
Greece.

[84] Papaphotis, G. and Tsaparlis, G. (2008). “Conceptual versus algorithmic

learning in high school chemistry: the case of basic quantum chemical
concepts, Part 2. Students’ common errors, misconceptions and difficul-
ties in understanding.” Chemistry Education: Research and Practice,
9(4), 332-340.

[85] Bethge, T., and Niedderer, H. (1996). “Students’ conceptions in quan-

tum physics.” Unpublished manuscript, http://134.102.186.148/pubs/
Niedderer/1995-AJP-TBHN.pdf.

[86] Galili, I., Goldberg, F. and Bendall, S. (1993). “Effects of prior knowledge
and instruction on understanding image formation.” Journal of Research
in Science Teaching, 30(3), 271-303.

[87] Galili, I. (1996). “Students’ conceptual change in geometrical optics.”

International Journal of Science Education, 18(7), 847-868.

[88] diSessa, A. (1998) “Knowledge in Pieces.” in Constructivism in Computer

Age. Hillsdale (NJ): Erlbaum.

[89] Minstrell, J. (1992). “Facets of students’ knowledge and relevant instruc-

tion.” In Research in Physics Learning: Theoretical Issues and Empirical
Studies. Kiel (DE): IPN (pp. 110-128).

[90] Grayson, D. J. (2004). “Concept substitution: A teaching strategy for

helping students disentangle related physics concepts.” American Journal
of Physics, 72(8), 1126-1133.

[91] Galili, I., and Hazan, A. (2000). “The influence of an historically oriented
course on students’ content knowledge in optics evaluated by means of
facets-schemes analysis.” American Journal of Physics, 68(S1), S3-S15.

[92] Guthrie, J. T. (1984). “Testing higher level skills.” Journal of Reading,

28, 188–190.

236
 [93] Frederiksen, J. R., and Collins, A. (1989). “A systems approach to edu-
cational testing.” Educational Researcher, 18(9), 27–32.

[94] Popham, W. J. (2006). Assessment for educational leaders. Boston (MA):

Pearson.

[95] Sebrechts, M. M., Bennett, R. E., and Rock, D. A. (1991). “Agreement

between expert-system and human raters’ scores on complex constructed-
response quantitative items.” Journal of Applied Psychology, 76(6), 856.

[96] Liu, O. L., Lee, H. S., and Linn, M. C. (2011). “Measuring knowledge
integration: Validation of four year assessments.” Journal of Research in
Science Teaching, 48(9), 1079-1107.

[97] Tuckman, B. (1994).“Tips for using essay tests.”Education Digest, 59(5),

71.

[98] Lee, H. S., Liu, O. L., and Linn, M. C. (2011). “Validating measurement
of knowledge integration in science using multiple-choice and explanation
items.” Applied Measurement in Education, 24(2), 115-136.

[99] Haslam, F., and Treagust, D. F. (1987). “Diagnosing secondary students’

misconceptions of photosynthesis and respiration in plants using a two-
tier multiple choice instrument.” Journal of Biological Education, 21(3),
203-211.

[100] Lee, H. S., Linn, M. C., Varma, K., and Liu, O. L. (2010). “How do
technology-enhanced inquiry science units impact classroom learning?.”
Journal of Research in Science Teaching, 47(1), 71-90.

[101] Linn, M. C. (2006). “The knowledge integration perspective on learning

and instruction.” In R. K. Sawyer (Ed.), The Cambridge handbook of
the learning sciences. New York (NY): Cambridge University Press (pp.
243–264).

[102] Lee, H. S., Linn, M. C., Varma, K., and Liu, O. L. (2010). “How do
technology-enhanced inquiry science units impact classroom learning?”
Journal of Research in Science Teaching, 47(1), 71-90.

[103] Chiu, J. L., and Linn, M. C. (2011). “Knowledge integration and WISE
engineering.” Journal of Pre-College Engineering Education Research,
1(1), 2.

[104] DeBoer, G. E., Lee, H. S., and Husic, F. (2008). “Assessing inte-
grated understanding of science.” In Designing coherent science edu-
cation: Implications for curriculum, instruction, and policy. New York
(NY): Columbia University (pp. 153-182).

237
 Bibliography

[105] Stone, E. M. (2014). “Guiding Students to Develop an Understanding of

Scientific Inquiry: A Science Skills Approach to Instruction and Assess-
ment.” CBE-Life Sciences Education, 13(1), 90-101.

[106] Chang, H. Y., Zhang, Z. H., and Chang, S. Y. (2014). “Adaptation of an

Inquiry Visualization Curriculum and its Impact on Chemistry Learn-
ing.” The Asia-Pacific Education Researcher, 23(3), 605-619.

[107] Svihla, V., and Linn, M. C. (2012). “A design-based approach to fostering

understanding of global climate change.” International Journal of Science
Education, 34(5), 651-676.

[108] Lee, H. S., and Liu, O. L. (2010). “Assessing learning progression of

energy concepts across middle school grades: The knowledge integration
perspective.” Science Education, 94(4), 665-688.

[109] Akarsu, B. (2011). “Instructional designs in quantum physics: A critical

review of research.” Asian J. Applied Sci., 4, 112-118.

[110] Falk, J., Linder, C., and Kung, R. L. (2007). “Review of empirical studies
into students’ depictions of quantum mechanics.” Educational Research
Review.

[111] Baily, C. (2011). “Perspectives in Quantum Physics: Epistemological,

Ontological and Pedagogical. An investigation into student and expert
perspectives on the physical interpretation of quantum mechanics, with
implications for modern physics instruction.” Doctoral Dissertation, Uni-
versity of Colorado.

[112] Escalada, L. T. (1997). “Investigating the applicability of activity-based

quantum mechanics in a few high school physics classrooms.” Doctoral
dissertation, Kansas State University.

[113] Ambrose, B. S. (1999). “Investigation of student understanding of the

wave-like properties of light and matter.” Doctoral dissertation, Univer-
sity of Washington.

[114] Falk, J. (2007). “Students’ depictions of quantum mechanics: a contem-

porary review and some implications for research and teaching.” Doctoral
dissertation, Uppsala University.

[115] Rao, S. K. (2012). “The Journey from Classical to Quantum Think-

ing: An Analysis of Student Understanding Through the Lens of Atomic
Spectra.” Doctoral dissertation, Berkeley University.

[116] Bertozzi, E. (2010). “Reconstructing Quantum Field Theory from an

Educational Perspective.” Doctoral dissertation, University of Bologna

238
[117] Telichevesky, L. (2015). “Uma perspectiva sociocultural para a introdu-
cao de conceitos de fı́sica quântica no ensino médio: análise das intera-
coes discursivas em uma unidade didática centrada no uso do interfer-
ômetro virtual de Mach-Zehnder.” Doctoral dissertation, University of
Rio Grande do Sul.
[118] Ireson, G. (2000). “The quantum understanding of pre-university physics
students.” Physics Education 35(1).
[119] Papaphotis, G., and Tsaparlis, G. (2008). “Conceptual versus algorithmic
learning in high school chemistry: the case of basic quantum chemical
concepts. Part 1. Statistical analysis of a quantitative study.” Chemistry
Education Research and Practice, 9(4), 323-331.
[120] Arons, A. B. (1990). A Guide to Introductory Physics Teaching. New
York (NY): Wiley and Sons.
[121] Wuttiprom, S., Sharma, M. D., Johnston, I. D., Chitaree, R. and
Soankwan, C. (2009). “Development and use of a conceptual survey in
introductory quantum physics,” International Journal of Science Educa-
tion, 31(5), 631–654.
[122] McKagan, S. B., Perkins, K. K., and Wieman, C. E. (2010). “Design
and validation of the quantum mechanics conceptual survey.” Physical
Review Special Topics-Physics Education Research, 6(2), 020121.
[123] McKagan, S. B. and Wieman, C. E. (2006). “Exploring student under-
standing of energy through the quantum mechanics conceptual survey.”
In Heron, P., McCullough, L. and Marx, J. (eds.) Physics Education
Research Conference Proceedings 2005.
[124] Olsen, R. V. (2002). “Introducing quantum mechanics in the upper sec-
ondary school: a study in Norway.” International Journal of Science Ed-
ucation, 24(6), 565-574.
[125] Baily, C., and Finkelstein, N. D. (2010).“Teaching and understanding
of quantum interpretations in modern physics courses.” Physical Review
Special Topics-Physics Education Research 6(1).
[126] Deslauriers, L. and Wieman, C. (2011). “Learning and retention of quan-
tum concepts with different teaching methods,” Physical Review Special
Topics-Physics Education Research, 7(1), 010101.
[127] Gil, D. and Solbes, J. (1993). “The introduction of modern physics: over-
coming a deformed vision of science,” International Journal of Science
Education, 15(3), 255-260.
[128] Mashhadi, A. (1995). “Students’ conceptions of quantum physics.” In
Thinking physics for teaching. New York (NY): Springer (pp. 313-328).

239
 Bibliography

[129] Mashhadi, A. and Woolnough, B. (1999). “Insights into students’ under-

standing of quantum physics: visualizing quantum entities.” European
Journal of Physics, 20(6), 511.

[130] Fischler, H. (1999) “Introduction to quantum physics. Development and

evaluation of a new course.” In Papers presented at the annual meet-
ing National Association for Research in Science Teaching. March, 1999
(p.32).

[131] Mannila, K., Koponen, I. T. Niskanen, J. A. (2002). “Building a picture

of students’ conceptions of wave-and particle-like properties of quantum
entities.” European Journal of Physics, 23(1), 45.

[132] Ambrose, B. S., Shaffer, P. S., Steinberg, R. N., and McDermott, L. C.

(1999). “An investigation of student understanding of single-slit diffrac-
tion and double-slit interference.” American Journal of Physics, 67(2),
146-155.

[133] Thacker, B. A. (2003). “A study of the nature of students’ models of

microscopic processes in the context of modern physics experiments.”
American Journal of Physics, 71(6), 599-606.

[134] Johnston, I. D., Crawford, K., and Fletcher, P. R. (1998). “Student diffi-
culties in learning quantum mechanics,” International Journal of Science
Education, 20(4), 427-446.

[135] Rebello, N. S., and Zollman, D. (1999). “Conceptual understanding of

quantum mechanics after using hands-on and visualization instructional
materials.” In Papers presented at the annual meeting National Associ-
ation for Research in Science Teaching March, 1999 (p. 2).

[136] Giliberti, M. (2006). Results from the IDIFO Master Workshop, Udine,
8.9.2006, reported in Fantini, P. and Levrini, O. “Fisica quantistica a
scuola: Indicazioni ministeriali, risultati di ricerca in didattica della fisica
e nuovi orizzonti” AIF seminar, Pavia, 30.10.2013.

[137] Müller, R. and Wiesner, H. (2002). “Teaching quantum mechanics on an

introductory level,” American Journal of Physics, 70(3), 200-209.

[138] Akarsu, B. (2010). “An extensive Study of Teaching/Learning Quantum

Mechanics in College.” arXiv preprint arXiv:1002.4975.

[139] McKagan, S. B., Handley, W., Perkins, K. K., and Wieman, C. E.

(2009). “A research-based curriculum for teaching the photoelectric ef-
fect.” American Journal of Physics, 77, 87–94.

[140] Ireson, G. (2000).“The quantum understanding of pre-university physics

students.” Physics Education, 35(1), 15.

240
[141] Kroemer, A. (1994). “Investigating quantum physics in the high schools
environments.” Physics Education, 56(3), 127-139.

[142] Vokos, S., Shaffer, P. S., Ambrose, B. S., and McDermott, L. C. (2000).
“Student understanding of the wave nature of matter: Diffraction and
interference of particles.” American Journal of Physics, 68(S1), S42-S51.

[143] Cheong, Y. W. and Song, J. (2013). “Different Levels of the Meaning of

Wave-Particle Duality and a Suspensive Perspective on the Interpreta-
tion of Quantum Theory.” Science and Education, 23(5), 1011–1030.

[144] Lautesse, P., Valls, A. V., Ferlin, F., Héraud, J. L., and Chabot,
H. (2015). “Teaching Quantum Physics in Upper Secondary School in
France.” Science and Education, 1-19.

[145] Fanaro, M., Otero, M. R., and Arlego, M. A. (2009). “Teaching the
foundations of quantum mechanics in secondary school: a proposed con-
ceptual structure.” Investigações em Ensino de Ciências, 14(1), 37-64.

[146] Bunge, M. (1973). Philosophy of physics. Dordrecht (NL): D. Reidel

Publishing Company.

[147] Bunge, M. (2003). “Twenty-five centuries of quantum physics: From

Pythagoras to us, and from subjectivism to realism.” Science and Ed-
ucation, 12(5–6), 445–466.

[148] Lévy-Leblond, J. M. (1977) “Towards a proper quantum theory.” In

Quantum Mechanics, a half century later (pp. 171-206). Dordrecth (NL):
Springer Netherland.

[149] Lévy-Leblond, J. M. (1999). “Quantum words for a quantum world.”

In Epistemological and Experimental Perspectives on Quantum Physics.
Dordrecht (NL): Springer Netherlands (pp. 75-87).

[150] Lévy-Leblond, J. M. (2003). “On the nature of quantons”. Science and

Education, 12(5), 495-502.

[151] de los Ángeles Fanaro, M., Arlego, M., and Otero, M. R. (2014). “The
double slit experience with light from the point of view of Feynman’s
sum of multiple paths.” Revista Brasileira de Ensino de Fısica, 36(2),
2308.

[152] Malgieri, M., Onorato, P., and De Ambrosis, A. (2014). “Teaching quan-
tum physics by the sum over paths approach and GeoGebra simulations.”
European Journal of Physics, 35(5), 055024.

[153] Malgieri, M., Onorato, P., and De Ambrosis, A. (2015) “What is light?
From optics to quantum physics through the sum over paths approach.”

241
 Bibliography

Proceedings of the GIREP-MPTL International Conference on Teach-

ing/Learning Physics: Integrating Research into Practice, July 7-12,
2014, University of Palermo, Italy.

[154] Malgieri, M., Onorato, P., and De Ambrosis, A. (2015). “Insegnare la

fisica quantistica a scuola: un percorso basato sul metodo dei cammini
di Feynman.” Giornale di fisica, 56(1).

[155] Malgieri, M., Onorato, P., and De Ambrosis, A. (2014) “Quantum me-
chanics and conceptual change: an educational proposal based on Feyn-
man’s sum over paths approach.”Talk at the 9th International Conference
on Conceptual Change, August 26-29 2014, Bologna.

[156] Euler, M., Hanselmann, M., Müller, A., and Zollman, D. (1999). “Stu-
dents’ views of models and concepts in modern physics”. In Papers pre-
sented at the annual meeting National Association for Research in Sci-
ence Teaching March, 1999 (p. 15).

[157] Griffiths, A. K., and Preston, K. R. (1992). “Grade-12 students’ miscon-

ceptions relating to fundamental characteristics of atoms and molecules.”
Journal of research in science teaching, 29(6), 611-628.

[158] Stavrou, D., Hadzidaki, P., and Kalkanis, G. (1999). “An instructional
model for a qualitative approach to QM concepts.” 8th Earli European
Conference, Göteborg, Sweden.

[159] Olsen, R. V. (2001). “A study of Norwegian upper secondary physics

specialists’ understanding of quantum physics.” In proceedings of the
3rd ESERA Conference, Thessaloniki, Greece.

[160] Nakiboglu, C. (2003). “Instructional misconceptions of Turkish prospec-

tive chemistry teachers about atomic orbitals and hybridization.” Chem-
istry Education Research and Practice, 4(2), 171-188.

[161] MacKinnon, G. R. (1999). “Students’ understanding of Orbitals: A Sur-

vey.” http://files.eric.ed.gov/fulltext/ED433248.pdf.

[162] Unlu, P. (2010). “Pre-service physics teachers’ ideas on size, visibility and
structure of the atom.” European Journal of Physics, 31(4), 881-892.

[163] Cros, D., and Chastrette, M. (1988). “Conceptions of second year univer-
sity students of some fundamental notions in chemistry.” International
Journal of Science Education, 10(3), 331-336.

[164] Ke, J. L., Monk, M. and Duschl, R. (2005). “Learning Introductory

Quantum Physics: Sensori-motor experiences and mental models,” In-
ternational Journal of Science Education, 27(13), 1571–1594.

242
[165] Budde, M., Niedderer, H., Scott, P., and Leach, J. (2002). “ ’Electro-
nium’: a quantum atomic teaching model.” Physics Education, 37(3),
197-203.

[166] Taber, K. S. (2001). “When the analogy breaks down: modelling the
atom on the solar system.” Physics Education, 36(3), 222-226.

[167] Hadzidaki, P., Kalkanis, G. and Stavrou, D. (2000). “Quantum mechan-

ics: a systemic component of the modern physics paradigm,” Physics
Education, 35(6), 386–392.

[168] Niedderer, H., Bethge, T., and Cassens, H. (1990).“A simplified quantum
model: a teaching approach and evaluation of understanding.” in Lijnse,
P.L., Licht, P. , de Vos, W. and Waarlo, A.J. (Eds.) Relating Macroscopic
Phenomena to Microscopic Particles: A Central Problem in Secundary
Science Education. Utrecht (NL): CD-β Press (pp. 67-80).

[169] Niedderer, H., Bethge, T., Cassens, H., and Petri, J. (1997). “Teaching
quantum atomic physics in college and research results about a learning
pathway.” In AIP conference proceedings (pp. 659-668). IOP Institute of
Physics.

[170] Petri, J., and Niedderer, H. (1998). “A learning pathway in high school
level quantum atomic physics.” International Journal of Science Educa-
tion, 20(9), 1075–1088.

[171] Budde, M., Niedderer, H., Scott, P., and Leach, J. (2002). “The quan-
tum atomic model ‘Electronium’: a successful teaching tool.” Physics
Education, 37(3), 204.

[172] Herrmann, F. (2000).”The Karlsruhe physics course.” European Journal

of Physics, 21(1), 49.

[173] de Broglie, L. “Recherches sur la théorie des quanta.” Doctoral disserta-

tion (1924), Faculté des Sciences, Paris.

[174] Madelung, E. (1926). “Quantum theory in hydrodynamical form.” Z.

Phys, 40, 332-336.

[175] Hood, C. G. (1993). “Teaching about quantum theory.” The Physics

Teacher, 31(5), 290- 293.

[176] Escalada, L. T., Rebello, N. S., and Zollman, D. A. (2004). “Student

explorations of quantum effects in LEDs and luminescent devices,” The
Physics Teacher, 42(3), 173-179.

[177] Taber, K. S. (2002). “Conceptualizing quanta: Illuminating the ground

state of student understanding of atomic orbitals.” Chemistry Education:
Research and Practice in Europe, 3(2), 145-158.

243
 Bibliography

[178] McKagan, S. B., Perkins, K. K., and Wieman, C. E. (2008). “Why we

should teach the Bohr model and how to teach it effectively.” Physical
Review Special Topics-Physics Education Research, 4(1), 010103.

[179] Maftei, G. (2011). “The angular momentum and the Bohr’s hydrogen
atom model-an approach to a more effective teaching.” Journal of edu-
cational sciences and psychology, 1(1).

[180] Shiland, T. W. (1997). “Quantum mechanics and conceptual change in

high school chemistry textbooks.” Journal of Research in Science Teach-
ing, 34(5), 535-545.

[181] McKagan, S. B., Perkins, K. K., Dubson, M., Malley, C., Reid, S.,
LeMaster, R., and Wieman, C. E. (2008). “Developing and researching
PhET simulations for teaching quantum mechanics.” American Journal
of Physics, 76(4), 406-417.

[182] http://phet.colorado.edu

[183] Zollman, D. A., Rebello, N. S., and Hogg, K. (2002). “Quantum me-
chanics for everyone: Hands-on activities integrated with technology,”
American Journal of Physics, 70(3), 252-259.

[184] http://phys.educ.ksu.edu/

[185] Jammer, M. (1974) The Philosophy of Quantum Mechanics. New York

(NY): Wiley.

[186] Cassidy, D.C. (1998) “Answer to the question: When did the indetermi-
nacy principle become the uncertainty principle?” American Journal of
Physics 66(4) 278-279.

[187] Sadaghiani, H. R. (2005). “Conceptual and mathematical barriers to

students learning quantum mechanics.” Doctoral dissertation, The Ohio
State University.

[188] Styer, D. F. (1996). “Common misconceptions regarding quantum me-

chanics.” American Journal of Physics, 64, 1202-1202.

[189] Fantini, P. and Levrini, O. (2013) “Fisica quantistica a scuola: Indi-

cazioni ministeriali, risultati di ricerca in didattica della fisica e nuovi
orizzonti” AIF seminar, Pavia, 30.10.2013.

[190] Müller, R., and Wiesner, H. (1999). “Students’ conceptions of quantum

physics.”. In Papers presented at the annual meeting National Associa-
tion for Research in Science Teaching March, 1999 (p. 20).

244
[191] Hadzidaki, P. (2008). “The Heisenberg Microscope: A Powerful Instruc-
tional Tool for Promoting Meta-Cognitive and Meta-Scientific Thinking
on Quantum Mechanics and the ’Nature of Science’.” Science and Edu-
cation, 17(6), 613-639.

[192] Velentzas, A., and Halkia, K. (2011). “The ’Heisenberg’s Microscope’ as

an example of using thought experiments in teaching physics theories to
students of the upper secondary school.” Research in Science Education,
41(4), 525-539.

[193] Cutnell J. D. and Johnson K. W. (2012). Fisica blu. Italian ed. tr. Romeni
C., Bologna (IT): Zanichelli.

[194] Cutnell J. D. and Johnson K. W. (2007) Physics. New York (NY): John
Willey and Sons.

[195] Halliday, D., Resnick, R., and Walker, J. (2010). Fundamentals of physics
extended. New York (NY): John Wiley and Sons.

[196] Johansson, K. E. and Milstead, D. (2008). “Uncertainty in the classroom–

teaching quantum physics,” Physics Education, 43(2), 173.

[197] Muiño, P. L. (2000). “Introducing the Uncertainty Principle Using

Diffraction of Light Waves.” Journal of Chemical Education, 77(8), 1025.

[198] Matteucci, G., Ferrari, L., and Migliori, A. (2010). “The Heisenberg
uncertainty principle demonstrated with an electron diffraction exper-
iment.” European Journal of Physics, 31(5), 1287.

[199] Dreyfus, B. W., Sohr, E. R., Gupta, A., and Elby, A. (2015). “ ’Classical-
ish’: Negotiating the boundary between classical and quantum particles.”
arXiv preprint arXiv:1507.00668.

[200] Morgan, J. T., Wittmann, M. C., and Thompson, J. R. (2004). “Student

understanding of tunneling in quantum mechanics: Examining interview
and survey results for clues to student reasoning.” In 2003 Physics Edu-
cation Research Conference (Vol. 720, pp. 97-100).

[201] Morgan, J. T., and Wittmann, M. C. (2006). “Examining the evolution

of student ideas about quantum tunneling.” 2005 Physics Education Re-
search Conference. Vol. 818.

[202] Wittmann, M. C., Morgan, J. T., and Bao, L. (2005). “Addressing stu-
dent models of energy loss in quantum tunnelling.” European Journal of
Physics, 26(6), 939-950.

[203] Morgan, J. T. (2006). “Investigating how Students Think About and

Learn Quantum Physics: An Example from Tunneling.” Doctoral disser-
tation, University of Maine.

245
 Bibliography

[204] Brookes, D. T., and Etkina, E. (2006).“Do our words really matter? Case
studies from quantum mechanics.” In 2005 Physics Education Research
Conference (Vol. 818, pp. 57-60).

[205] Koopman, L., and Ellermeijer, T. (2005). “Understanding student diffi-

culties in first year quantum mechanics courses.” In Proceedings of the
EPEC-1 2005 Conference (pp. 4-8).

[206] Domert, D., Linder, C., and Ingerman, A. (2005). “Probability as a con-
ceptual hurdle to understanding onedimensional quantum scattering and
tunnelling.” European Journal of Physics, 26(1), 47-59.

[207] Redish, E. F., Wittmann, M. C., and Steinberg, R. N. (2000). “Affect-

ing Student Reasoning in the Context of Quantum Tunneling.” AAPT
summer meeting 2000. Guelph, Ontario.

[208] Bao, L., and Redish, E. F. (2002). “Understanding probabilistic interpre-

tations of physical systems: A prerequisite to learning quantum physics.”
American Journal of Physics, 70(3), 210-217.

[209] Morgan, M. J., and Jakovidis, G. (1994). “Characteristic Energy Scales

of Quantum Systems.” Physics Teacher, 32(6), 354-58.

[210] Singh, C. (2001). “Student understanding of quantum mechanics.” Amer-

ican Journal of Physics, 69(8), 885-895.

[211] Zhu, G., and Singh, C. (2012). “Improving students’ understanding of

quantum measurement. I. Investigation of difficulties.” Physical Review
Special Topics-Physics Education Research, 8(1), 010117.

[212] Singh, C. (2008). “Student understanding of quantum mechanics at the

beginning of graduate instruction.” American Journal of Physics, 76(3),
277-287.

[213] Zhu, G., and Singh, C. (2012). “Surveying students’ understanding of

quantum mechanics in one spatial dimension.” American Journal of
Physics, 80(3), 252-259.

[214] Wosilait, K., Heron, P. R., Shaffer, P. S., and McDermott, L. C. (1999).
“Addressing student difficulties in applying a wave model to the inter-
ference and diffraction of light.” American Journal of Physics, 67(S1),
S5-S15.

[215] Colin, P., and Viennot, L. (2001). “Using two models in optics: Students’
difficulties and suggestions for teaching.” American Journal of Physics,
69(S1), S36-S44.

246
[216] Scherr, R. E., Close, H. G., McKagan, S. B., and Vokos, S. (2012). “Rep-
resenting energy. I. Representing a substance ontology for energy.” Phys-
ical Review Special Topics-Physics Education Research, 8(2), 020114.
[217] Rinaudo, G. (2013)“Il quanto di luce e la fisica quantistica.” Talk at the
AIF school on quantum physics for teachers, 6th November, 2013.
[218] Heron, P., Michelini, M., and Stefanel, A. (2008). “Teaching and learn-
ing the concept of energy in primary school.” In Resúmenes de la XIII
Conferencia Internacional GIREP. Nicosia, Cipre (Vol. 96).
[219] Duit, R. (1987). “Should energy be illustrated as something quasi-
material?” International Journal of Science Education, 9(2), 139-145.
[220] Papadouris, N., Constantinou, C. P., and Kyratsi, T. (2008). Students’
use of the energy model to account for changes in physical systems.
Journal of Research in science teaching, 45(4), 444-469.
[221] Doménech, J. L., Gil-Pérez, D., Gras-Martı́, A., Guisasola, J., Martı́nez-
Torregrosa, J., Salinas, J., Trumpers, R., Valdé, P. and Vilches, A.
(2007). “Teaching of energy issues: A debate proposal for a global re-
orientation.” Science and Education, 16(1), 43-64.
[222] Sadaghiani, H., and Bao, L. (2006). “Student difficulties in understanding
probability in quantum mechanics.” In 2005 Physics Education Research
Conference (Vol. 818, pp. 61-64).
[223] Bao, L., Redish, E. F., and Steinberg, R. N. (1998). “Student misun-
derstandings of the quantum wavefunction.” AAPT Announcer, 28(2),
92.
[224] Mach, E. (1898). The Science of Mechanics; a Critical and Historical
Account of its Development. Chicago (IL): Open Court, 1960.
[225] Matthews, M.R. (1994), Science Teaching: The Role of History and Phi-
losophy of Science, New York (NY): Routledge.
[226] Galili, I., and Hazan, A. (2001). The effect of a history-based course in
optics on students’ views about science. Science and Education, 10(1-2),
7-32.
[227] Born, M., Blin-Stoyle, R. J., and Radcliffe, J. M. (1989). Atomic physics.
New York (NY): Dover Publications.
[228] Messiah, A., (1961) Quantum mechanics, vol. I Amsterdam (NL): North-
Holland.
[229] Longair, M. (2013). textslQuantum concepts in physics: an alternative
approach to the understanding of quantum mechanics. New York (NY):
Cambridge University Press.

247
 Bibliography

[230] Sperandeo R. M. (2004). “The Pre-Service Teacher Education model

implemented by the FFC research project involving 8 italian universi-
ties: guidelines and preliminary results, in Quality Development in the
Teacher Education and Training.” in Michelini M. ed., Quality Develop-
ment in the Teacher Education and Training, Udine: Girep-Forum.

[231] Nashon, S., Nielsen, W., and Petrina, S. (2008). “Whatever happened to
STS? Pre-service physics teachers and the history of quantum mechan-
ics.” Science and Education, 17(4), 387-401.

[232] Gil, D., and Solbes, J. (1993). “The introduction of modern physics:
overcoming a deformed vision of science.”International Journal of Science
Education, 15(3), 255-260.

[233] Townsend, J. S. (2000) A modern approach to quantum mechanics.

Sausalito (CA): University Science Books.

[234] Michelini, M., Ragazzon, R., Santi, L., and Stefanel, A. (2000). “Proposal
for quantum physics in secondary school.” Physics Education, 35(6), 406.

[235] Pospiech, G. (2000). “Uncertainty and complementarity: the heart of

quantum physics.” Physics education, 35(6), 393.

[236] Close, H., Schiber, C., Close, E. and Donnelly, D. (2013). “Students’
dynamic geometric reasoning about quantum spin-1/2 states,” PERC
2013 Proceedings, 93-96.

[237] Tarozzi, F. (2005) “Un progetto di insegnamento della meccanica quantis-

tica a livello di scuola secondaria superiore: alla ricerca di un formalismo
possibile.” M.S. thesis, University of Bologna.

[238] Pospiech, G., Michelini, M., Stefanel, A., and Santi, L. (2008). “Central
features of quantum theory in physics education.” In Discussion Work-
shop B, GIREP 2007 Proceedings (pp. 85-87).

[239] Michelini M., Ragazzon R., Santi L., Stefanel A. (2004). “Discussion of
a didactic proposal on quantum mechanics with secondary school stu-
dents”, Il Nuovo Cimento, 27 C, 5, 555-567.

[240] Michelini, M. , and Stefanel, A. . (2007). “Interpreting Diffraction Using

the Quantum Model.” Proceedings GIREP Conference2006: Modeling in
Physics and Physics Education, 711-715.

[241] Ghirardi G.C., Grassi R., and Michelini M. (1995). “A Fundamental Con-
cept in Quantum Theory: The Superposition Principle”, in C. Bernardini
et a.l (eds.) Thinking Physics for Teaching , Plenum: Aster, pp. 329-334.

248
[242] Ghirardi G.C., Grassi R., Michelini M. (1997). “Introduzione delle idee
della fisica quantistica e il ruolo del principio di sovrapposizione lineare,”
La Fisica nella Scuola, XXX, 3 Sup., Q7, 46-57.

[243] Schneider, M. B., and LaPuma, I. A. (2002). “A simple experiment for

discussion of quantum interference and which-way measurement.” Amer-
ican Journal of Physics, 70(3), 266-271.

[244] Pospiech, G. (1999). “Teaching the EPR paradox at high school?”

Physics Education, 34(5), 311.

[245] Pospiech, G. (2001). “Experiences with a modern course in quantum

physics.” In Research in Science Education-Past, Present, and Future
(pp. 89-94). Springer Netherlands.

[246] Sakurai, J.J. (1985), Modern Quantum Mechanics. Reading (MA):

Addison-Wesley.

[247] Pospiech, G. . (1999). “A modern course in quantum physics for teacher

education.” Proceedings of second international ESERA conference (pp.
244-249).

[248] Michelini, M., Santi, L., and Stefanel, A. (2010). “Gli insegnanti ri-
flettono sui nodi concettuali della meccanica quantistica.” In Michelini,
M. (ed) Progetto IDIFO. Fisica moderna per la scuola. Materiali, as-
petti e proposte per l’innovazione didattica e l’orientamento. Udine (IT):
LithoStampa.

[249] Michelini, M., Santi, L., and Stefanel, A. (2010)“Impostazione alla Dirac.
La proposta didattica di Udine per la fisica quantistica.” In Formazione
a distanza degli insegnanti all’innovazione didattica in fisica moderna e
orientamento. (p. 13).

[250] Pospiech, G., and Schoene, M. (2012). “Quantenphysik in Schule

und Hochschule.” PhyDid B-Didaktik der Physik-Beiträge zur DPG-
Frühjahrstagung.

[251] Sadaghiani, H. (2014) “Investigations of spin first instructional approach

in teaching quantum mechanics,“ poster contribution, PERC 2014 Con-
ference.

[252] Hobson A. (2005), “Electrons as field quanta: A better way to teach

quantum physics in introductory general physics courses”, American
Journa of Physics, 73 (7).

[253] Hobson, A. (2013). “There are no particles, there are only fields.” Amer-
ican Journal of Physics, 81, 211.

249
 Bibliography

[254] Giliberti, M. (2002). “A modern teaching for modern physics in pre-

service teacher training.” Developing Formal, 400.

[255] Giliberti M., Lanz L., Cazzaniga L. (2002), “Quanta-MI: a modern teach-
ing for modern physics in preservice teachers training,” Paper presented
at the international GIREP Conference, Physics in new fields and mod-
ern applications - opportunities for physics education, August 5 - 9, 2002
Lund, Sweden.

[256] Bertozzi, E. (2010). “Hunting the ghosts of a ’strictly quantum field’: the
Klein–Gordon equation.” European Journal of Physics, 31(6), 1499.

[257] Bertozzi, E., Ercolessi, E., and Levrini, O. (2013). “Words and formu-
las in quantum field theory: Disentangling and reassembling the basic
concepts for teaching.” Physics Essays, 26(3).

[258] Giliberti, M., Lanz, L., and Cazzaniga, L. (2004). “Teaching quan-
tum physics to student teachers of SILSIS-MI.” In Second international
GIREP seminar “quality development in teacher education and training”,
selected contributions FORUM Editrice Universitaria Udinese, Udine
(pp. 425-429).

[259] Sciamanda, R. J. (2013). “There are no Particles, and There are no

Fields.” American Journal of Physics, 81(9), 645-645.

[260] Daniel, M. (2006). “Particles, Feynman diagrams and all that.” Physics
education, 41(2), 119.

[261] Van Den Berg, E., and Hoekzema, D. (2006). “Teaching conservation
laws, symmetries and elementary particles with fast feedback.” Physics
education, 41(1), 47.

[262] Hoekzema, D., Schooten, G., van den Berg, E., and Lijnse, P. (2005).
“Conservation laws, symmetries, and elementary particles.” The Physics
Teacher, 43(5), 266-271.

[263] Tarsitani, C. (2010) “Una proposta per l’insegnamento della fisica quan-
tistica. Formazione a distanza degli insegnanti all’innovazione didattica.”
in Fisica moderna e orientamento, 49.

[264] Giannelli A. and Tarsitani C. (2003) “Un progetto di introduzione alla

meccanica quantistica per i laureati in matematica,” La Fisica nella
Scuola, XXXVI, n. 3, pp. 103-114.

[265] Giannelli A. and Tarsitani C. (2004) “Teaching Quantum Mechanics to

future school teachers,” in Quality Development in Teacher Education
and Training. Selected contributions of the Second International GIREP
Seminar 2003, Forum, Udine, pp. 441-445.

250
[266] Bunge, M. (1967). Foundations of physics. New York (NY): Springer.

[267] Lévy-Leblond, J. M., and Balibar F. (1984) Quantique. Rudiments.

Paris(FR): InterEditions.

[268] Strauss, M. (1969). “Corrections to Bunge’s Foundations of Physics”

(1967). Synthese, 19(3), 433-442.

[269] Lévy-Leblond, J. M. (2004) “Les révolutions du XXe siècle”, cours au

Collége de la Cité des sciences, cours du jeudi 08 janvier 2004, “La
révolution quantique : une matiere qui défie l’intuition”, http://www.
fabriquedesens.net/Les-revolutions-du-XXe-siecle.

[270] Ryder, J. (2002). “School science education for citizenship: strategies

for teaching about the epistemology of science.” Journal of Curriculum
Studies, 34(6), 637-658.

[271] Galili, I. (2012). “Cultural content knowledge – The case of physics edu-
cation.” International Journal of Innovation in Science and Mathematics
Education 20(2).

[272] Galili, I. (2012). “Promotion of Content Cultural Knowledge through the

use of the History and Philosophy of Science”, Science and Education,
21(9), 1283-1316.

[273] Shulman, L. (1986). “Those who understand: Knowledge growth in

teaching.” Educational Researcher, 15(2), 4–14.

[274] Garritz, A. (2013). “Teaching the philosophical interpretations of quan-

tum mechanics and quantum chemistry through controversies.” Science
and Education, 22(7), 1787-1807.

[275] Braga, M., Guerra, A., and Reis, J. C. (2012). “The role of historical-
philosophical controversies in teaching sciences: The debate between biot
and ampere.” Science and Education, 21(6), 921-934.

[276] Niaz, M. (2009). “Progressive transitions in chemistry teachers’ under-

standing of nature of science based on historical controversies.” Science
and Education, 18(1), 43-65.

[277] Fantini, P., Levrini, O., and Grimellini Tomasini, N. (2005).

“L’irriducibile complessità del pensiero scientifico: ostacolo o sfida per
la diffusione della cultura scientifica?” In Enseñanza de las Ciencias (pp.
1-5).

[278] Schlosshauer, M. (2005). “Decoherence, the measurement problem, and

interpretations of quantum mechanics,” Reviews of Modern Physics,
76(4), 1267.

251
 Bibliography

[279] Schlosshauer, M., Kofler, J., and Zeilinger, A. (2013). “A snapshot

of foundational attitudes toward quantum mechanics.” arXiv preprint
arXiv:1301.1069.

[280] Sommer, C. (2013). “Another Survey of Foundational Attitudes Towards

Quantum Mechanics.” arXiv preprint arXiv:1303.2719.

[281] Dubson, M., Goldhaber, S., Pollock, S., and Perkins, K. (2009). “Faculty
disagreement about the teaching of quantum mechanics.” In AIP Conf.
Proc (Vol. 1179, p. 137).

[282] Baily, C., and Finkelstein, N. D. (2010). “Refined characterization of stu-

dent perspectives on quantum physics.” Physical Review Special Topics-
Physics Education Research, 6(2), 020113.

[283] Stapp, H. P. (1972). “The Copenhagen Interpretation.” American Journal

of Physics, 40(8), 1098-1116.

[284] Ballentine, L. (1970).“The Statistical Interpretation of Quantum Me-

chanics.” Rev. Mod. Phys. 42, 358-380.

[285] Pessoa Jr, O. (1997). “Interferometria, Interpretacao e Intuicao: uma

Introducao Conceitual á Fisica Quântica.” Revista Brasileira de Ensino
de Fisica, 19(1).

[286] Pereira, A., Ostermann, F. and Cavalcanti, C. (2009). “On the use of
a virtual Mach–Zehnder interferometer in the teaching of quantum me-
chanics.” Physics Education, 44(3), 281.

[287] Ostermann, F., and Prado, S. D. (2005). “Conceptual and Epistemolog-

ical discussions on Quantum Mechanics in a Virtual Laboratory.” arXiv
preprint physics/0507064.

[288] Kohnle, A., Bozhinova, I., Browne, D., Everitt, M., Fomins, A., Kok,
P., and Swinbank, E. (2014). “A new introductory quantum mechanics
curriculum.” European Journal of Physics, 35(1), 015001.

[289] de los Ángeles Fanaro, M., and Otero, M. R. (2008). “Basics Quantum
Mechanics teaching in secondary school: One conceptual structure based
on Paths Integrals Method.” Latin-American Journal of Physics Educa-
tion, 2(2), 3.

[290] Ostermann, F., Cavalcanti, C. J. H., Prado, S. D., and Ricci, T. F.

(2009). “Fundamentos da fı́sica quântica à luz de um interferômetro vir-
tual de Mach-Zehnder”. Revista Electrónica de Ensenanza de las Cien-
cias, 8(3), 1094-1116.

[291] Hobson, A. (2012). “Teaching Quantum Nonlocality.” The Physics

Teacher, 50(5), 270-273.

252
[292] Hobson, A. (1996). Teaching quantum theory in the introductory course.
Physics Teacher, 34, 202-209.

[293] Holbrow, C. H., Amato, J. C., Galvez, E. J., and Lloyd, J. N. (1995).
“Modernizing introductory physics.” American Journal of Physics,
63(12), 1078-1090.

[294] Holbrow, C. H., Lloyd, J. N., Amato, J. C., Galvez, E., and Parks, M. E.
(2010). Modern introductory physics. New York (NY): Springer Science
and Business Media.

[295] Grangier, P., Roger, G. and Aspect, A. (1986). “Experimental evidence

for a photon anticorrelation effect on a beam splitter: a new light on
single-photon interferences.” EPL (Europhysics Letters), 1(4), 173.

[296] Hong, C. K., Ou, Z. Y., and Mandel, L. (1987). “Measurement of subpi-
cosecond time intervals between two photons by interference.” Physical
Review Letters, 59(18), 2044.

[297] Rutten, N., van Joolingen, W. R. and van der Veen, J. T. (2012). “The
learning effects of computer simulations in science education.” Computers
and Education, 58(1), 136–153.

[298] Swaak, J. and De Jong, T. (2001). “Discovery simulations and the assess-
ment of intuitive knowledge.” Journal of Computer Assisted Learning,
17(3), 284–294.

[299] Blake, C. and Scanlon, E. (2007). “Reconsidering simulations in science

education at a distance: features of effective use.” Journal of Computer
Assisted Learning, 23(6), 491–502.

[300] Fraser, D., Allison, S., Coombes, H., Case, J. and Linder, C. (2006). “Us-
ing variation to enhance learning in engineering.” International Journal
of Engineering Education, 22(1), 102.

[301] Podolefsky, N. S., Perkins K. K. and Adams W. K. (2010) “Factors pro-

moting engaged exploration with computer simulations.” Phys. Rev. ST
Physics Ed. Research 6 020117, 1–11.

[302] Kohnle, A., Cassettari, D., Edwards, T. J., Ferguson, C., Gillies, A. D.,
Hooley, C. A., and Sinclair, B. D. (2012). “A new multimedia resource
for teaching quantum mechanics concepts.” American Journal of Physics,
80(2), 148-153.

[303] Kohnle, A. (2014). “Research-based interactive simulations to support

quantum mechanics learning and teaching.”, GIREP-MPTL 2014 Inter-
national Conference, 7-12 July 2014, Palermo.

[304] https://www.st-andrews.ac.uk/physics/quvis/

253
 Bibliography

[305] Belloni, M., Christian, W., and Cox, A. J. (2007). “Teaching qualitative
energy-eigenfunction shape with Physlets.” The Physics Teacher, 45(8),
488-491.
[306] http://www.compadre.org/pqp/
[307] Mason, B., Debowska,
, E., Arpornthip, T., Girwidz, R., Greczylo, T.,
Kohnle, A., Melder, T., Michelini, M., Santi, L. and Silva, J. (2014)
“Report and Recommendations on Multimedia Materials for Teaching
and Learning Quantum Physics.”http://www.um.es/fem/PersonalWiki/
uploads/MPTL/MMReport-2014-Quantum.pdf
[308] http://www.iapht.unito.it/qm/
[309] http://rcl-munich.informatik.unibw-muenchen.de/
[310] http://www.ises.info/index.php/en/laboratory
[311] https://wikis.mit.edu/confluence/display/ILAB2/iLabs
[312] http://www.didaktik.physik.uni-erlangen.de/quantumlab/
[313] Schulman, L. S. (1981). Techniques and applications of path integration.
New York (NY): John Wiley and Sons.
[314] Rattazzi, R. (2009) “The Path Integral Approach to Quantum Mechan-
ics”. Lecture Notes for Quantum Mechanics.
[315] Egli, C. (2004). “Feynman Path Integrals in Quantum Mechanics.” Lec-
ture Notes.
[316] Swanson, M. S. (1992). Path integrals and quantum processes. Boston
(MA): Academic Press Inc.
[317] Thaller, B. (2005). Advanced visual quantum mechanics. New York
(NY): Springer Science and Business Media.
[318] Born, M. and Wolf, E. (1959). Principles of optics. London (UK): Perg-
amon Press.
[319] Schwabl, F. (2005). Advanced quantum mechanics. New York (NY):
Springer Science and Business Media.
[320] Landau, L. D., and Lifshits, E. M. (1965) Quantum Mechanics: non-
Relativistic Theory. London (UK): Pergamon Press.
[321] Duffy, D. G. (2001). Green’s functions with applications. London (UK):
Chapman and Hall.
[322] Fried, H. M., and Muller, B. (Eds.) (2012). Vacuum structure in intense
fields (Vol. 255). New York (NY): Springer Science and Business Media.

254
[323] Sawant, R., Samuel, J., Sinha, A., Sinha, S., and Sinha, U. (2014). “Non-
classical paths in quantum interference experiments.” Physical Review
Letters 113(12).

[324] Gitin, A. V. (2013). “Huygens-–Feynman-–Fresnel principle as the basis

of applied optics.” Applied optics 52.31: 7419-7434.

[325] Jones, E. R., Bach, R. A., and Batelaan, H. (2015).“Path integrals, mat-
ter waves, and the double slit.” European Journal of Physics, 36(6),
065048

[326] Kleinert, H. (1996). Particles and Quantum Fields. Lecture Notes. http:
//www.physik.fu-berlin.de/kleinert/lecture-notes.html.

[327] Styer, D. F. “The Klein-Gordon Propagator” (1999) http://www.oberlin.

edu/physics/dstyer/StrangeQM/Klein-Gordon.pdf

[328] Gray, C. G. and Taylor, E. F. (2007). “When action is not least.” Amer-
ican Journal of Physics 75(5), 434–458.

[329] http://www.eftaylor.com/

[330] http://sciencesoftware.com.cn

[331] http://fdslive.oup.com/www.oup.com/oxed/secondary/science/
advancingphysics/index.html

[332] Ogborn, J. (1974). “Introducing quantum physics.” Physics Education,

9(7), 436.

[333] Ogborn J., Hanc J. and Taylor E. F. (2006) “Action on stage: Historical
introduction.” The Girep Conference “Modelling in Physics and Physics
Education” 20-25 August 2006, AMSTEL institute, Faculty of Science,
University of Amsterdam, Netherlands.

[334] http://modellus.fct.unl.pt/

[335] Ogborn, J. “Advancing Physics evaluated”. Physics education, 38(4), 330

(2003).

[336] Hanc, J. and Taylor, E. F. (2004). “From conservation of energy to the

principle of least action: a story line.” American Journal of Physics,
72(4), 514-521.

[337] Fabri, E. (1996) “Come introdurre la fisica moderna nella scuola secon-
daria superiore” LFNS 29, Suppl.n.1, p.63-80.

255
 Bibliography

[338] Borello L., Cuppari A., Greco M., Rinaudo G. and Rovero G. (2002) “Il
metodo della somma sui molti cammini di Feynman per l’introduzione
della Meccanica Quantistica” La Fisica nella Scuola, Suppl.n.2, p.119-
124.

[339] http://www.geogebra.org/

[340] Malgieri, M., Onorato, P., Da Ambrosis A. (2015) “A sum over paths ap-
proach to one-dimensional time independent quantum systems.” Ameri-
can Journal of Physics (accepted).

[341] Gutzwiller, M. C. (1971). “Periodic orbits and classical quantization con-

ditions.” Journal of Mathematical Physics, 12(3), 343-358.

[342] Gutzwiller, M. C. (1967) “Phase Integral Approximation in Momentum

Space and the Bound States of an Atom.” Journal of Mathematical
Physics, 8(10), 1979-2000.

[343] De Aguiar, M. A. M. (1993). “Exact Green’s function for the step and
square-barrier potentials.” Physical Review A, 48(4), 2567.

[344] Andrade, F. M. (2014). “Exact Green’s function for rectangular po-

tentials and its application to quasi-bound states.” Physics Letters A
378(21),1461-1468.

[345] Bransden, B. H. and Joachain, C. J. (2000). Quantum mechanics, Harlow

(UK): Pearson Education.

[346] Aronstein, D. L., and Stroud Jr, C. R. (2000). “General series solution for
finite square-well energy levels for use in wave-packet studies.” American
Journal of Physics, 68(10), 943-949.

[347] Gilbert, L. P., Belloni, M., Doncheski, M. A. and Robinett, R. W. (2005).

“More on the asymmetric infinite square well: energy eigenstates with
zero curvature.” European Journal of Physics, 26(5), 815.

[348] Blumel, R. (2011). Advanced Quantum Mechanics: The Classical-

Quantum Connection. Sudbury (MA): Jones and Bartlett Publishers.

[349] Jacobsson, R. (1966) “V. Light Reflection from Films of Continuously

Varying Refractive Index.” Progress in optics, 5, 247-286.

[350] Diamant, R. and Fernández-Guasti, M. (2009). “Light propagation in

1D inhomogeneous deterministic media: the effect of discontinuities.”
Journal of Optics A: Pure and Applied Optics 11(4), 045712.

[351] Aspect, A., and Grangier, P. (1987). “Wave-particle duality for single
photons.” Hyperfine Interactions, 37(1-4), 1-17.

256
[352] Mittelstaedt, P., Prieur, A., and Schieder, R. (1987). “Unsharp particle-
wave duality in a photon split-beam experiment.”Foundations of Physics,
17(9), 891-903.

[353] Grangier, P., Levenson, J. A., and Poizat, J. P. (1998). “Quantum non-
demolition measurements in optics.” Nature, 396(6711), 537-542.

[354] Zhou X. Y., Wang L. J. , and Mandel L. (1991). “Induced coherence

and indistinguishability in optical interference.” Physical Review Letters,
67(3), 318-321.

[355] Aspect, A., Dalibard, J., and Roger, G. (1982). “Experimental test of
Bell’s inequalities using time-varying analyzers.” Physical review letters,
49(25), 1804.

[356] Jennewein, T., Weihs, G., Pan, J. W., and Zeilinger, A. (2001). “Exper-
imental nonlocality proof of quantum teleportation and entanglement
swapping.” Physical review letters, 88(1), 017903.

[357] Fuchs, C. A. (2003). “Quantum mechanics as quantum information,

mostly.” Journal of Modern Optics, 50(6-7), 987-1023.

[358] Brukner, C. and Zeilinger, A. (2003) “Information and fundamental el-

ements of the structure of quantum theory,” In Time, Quantum and
Information Heidelberg (DE): Springer Berlin (pp. 323-354).

[359] Lim, Y. L., and Beige, A. (2005). “Generalized Hong–Ou–Mandel exper-

iments with bosons and fermions.” New Journal of Physics, 7(1), 155.

[360] Redish, E. F., and Hammer, D. (2009). “Reinventing college physics for
biologists: Explicating an epistemological curriculum.” American journal
of physics, 77(7), 629-642.

[361] Redish, E. F. (2010). “Introducing students to the culture of physics:

Explicating elements of the hidden curriculum.” American Institute of
Physics Conference Series (Vol. 1289, pp. 49-52).

[362] Blume-Kohout, R., and Zurek, W. H. (2006). “Quantum Darwinism:

Entanglement, branches, and the emergent classicality of redundantly
stored quantum information,” Physical Review A, 73(6), 062310.

[363] Zurek, W. H. (2009). “Quantum darwinism,” Nature Physics, 5(3), 181-

188.

[364] Nersessian, N. (1984) Faraday to Einstein: Constructing meaning in sci-

entific theories. Dordrecth (NL): Martinus Nijhoff Publishers.

[365] Peres, A. (1993). Quantum Theory: Concepts and Methods Dordrecth

(NL): Kluwer Academic Publishing.

257
 Bibliography

[366] Maccone, L. (2006). “Information-disturbance tradeoff in quantum mea-

surements.” Physical Review A, 73(4), 042307.

[367] D’Ariano, G. M. (2003). “On the Heisenberg principle, namely on

the information-disturbance trade-off in a quantum measurement.”
Fortschritte der Physik, 51(4-5), 318-330.

[368] Robertson, H. P. (1934). “An indeterminacy relation for several observ-

ables and its classical interpretation.” Physical Review, 46(9), 794.

[369] Clauser, J. F. (1974). “Experimental distinction between the quantum

and classical field-theoretic predictions for the photoelectric effect.”Phys-
ical Review D, 9(4), 853.

[370] Chinn, C. A., and Brewer, W. F. (1993). “The role of anomalous data
in knowledge acquisition: A theoretical framework and implications for
science instruction,” Review of educational research, 63(1), 1-49.

[371] Davies, P. (1985) Superforce. New York (NY): Simon and Schuster
(p.24).

[372] Nieves, L., Spavieri, G., Fernandez, B., and Guevara, R. A. (1997). “Mea-
suring the Planck constant with LED’s,” The Physics Teacher, 35(2),
108-109.

[373] Rute de Amorim, M., Sá Ferreira A., de Brito André P. S., (2014).
“Classroom fundamentals: measuring the Planck constant,” Science in
school, 28. http://www.scienceinschool.org/2014/issue28/planck

[374] Amrani, D. (2014). “Hydrogen Balmer series measurements and deter-

mination of Rydberg’s constant using two different spectrometers.” Eu-
ropean Journal of Physics, 35(4) 045001.

[375] Derbes, D. (1996). “Feynman’s derivation of the Schrödinger equation.”

American Journal of Physics, 64(7), 881–884.

[376] Taylor, G. I. (1909). “Interference Fringes with Feeble Light,” Proc. Cam.
Phil. Soc. 15, 114.

[377] http://www.physique.ens-cachan.fr/old/franges photon/interference.

htm

[378] Davisson, C. and Germer, L. H. (1927). “Diffraction of electrons by a

crystal of nickel.” Physical Review, 30(6), 705.

[379] Tonomura, A., Endo, J., Matsuda, T., Kawasaki, T. and Ezawa, H.
(1989). “Demonstration of single-electron buildup of an interference pat-
tern.” American Journal of Physics, 57, 117.

258
[380] Missiroli, G. F., Pozzi, G. and Valdre, U. (1981).“Electron interferometry
and interference electron microscopy.” Journal of Physics E: Scientific
Instruments, 14(6), 649.
[381] Zeilinger, A., Shull, C. G., Treimer, W. and Mampe, W. (1988). “Single-
and double-slit diffraction of neutrons.” Reviews of modern physics,
60(4), 1067-1073.
[382] Arndt, M., Nairz, O., Vos-Andreae, J., Keller, C., Van der Zouw, G. and
Zeilinger, A. (1999). “Wave–particle duality of C60 molecules.” Nature,
401(6754), 680-682.
[383] Goldstein, H., Poole, C. and Safko, J. (1980) Classical Mechanics. Read-
ing (MA): Addison-Wesley.
[384] http://www.pasco.com/support/technical-support/technote/
techIDlookup.cfm?TechNoteID=303
[385] http://physicsworld.com/cws/article/news/52707
[386] Padmanabhan, T. (2009) Quantum themes: the charms of the mi-
croworld. Singapore: World Scientific.
[387] Hagelstein, P. L., Senturia, S. D. and Orlando, T. P. (2004). Introductory
applied quantum and statistical mechanics. New York (NY): John Wiley
and Sons,
[388] Becchi, C. M., and D’Elia M. (2010). Introduction to the basic concepts
of modern physics. New York (NY): Springer Science and Business Me-
dia.
[389] Gamow, G. (1928). “The quantum theory of nuclear disintegration.” Na-
ture 122, 805-806.
[390] Winful, H. G. (2006). “Tunneling time, the Hartman effect, and super-
luminality: A proposed resolution of an old paradox.” Physics Reports,
436(1), 1-69.
[391] Duru, I. H., and Kleinert, H. (1982). “Quantum Mechanics of H-Atom
from Path Integrals.” Fortschritte der Physik, 30(8), 401-435.
[392] Berry, M. V. and Tabor, M. (1976). “Closed orbits and the regular bound
spectrum.” Proceedings of the Royal Society of London. A. Mathematical
and Physical Sciences, 349(1656), 101-123.
[393] Zeilinger, A. (1981). “General properties of lossless beam splitters in
interferometry.” American Journal of Physics 49.9, 882-883.
[394] Zetie, K. P., Adams, S. F. and Tocknell, R. M. (2000). “How does a
Mach-Zehnder interferometer work?” Physics Education 35(1), 46.

259
 Bibliography

[395] Lee, H., Kok, P. and Dowling, J. P. (2002). “A quantum Rosetta stone
for interferometry.” Journal of Modern Optics,” 49(14-15), 2325-2338.

[396] Levrini, O., Fantini, P., Pecori, B., Gagliardi, M., Scarongella, M., and
Tasquier, G. (2010). “A longitudinal approach to appropriation of science
ideas: a study of students’ trajectories in thermodynamics,” In Proceed-
ings of the 9th International Conference of the Learning Sciences-Volume
1. International Society of the Learning Sciences (pp. 572-579).

[397] Giordano, E. (2011). “Imparare sperimentando,” Psicologia

dell’Educazione, 5, 177-192.

[398] Maxwell, J. C. (1864) “On Faraday’s lines of force,” Transactions of the

Cambridge Philosophical Society, 10, 27.

[399] Hertz, H. (1900) The principles of mechanics presented in a new form.

Mineola (NY): Dover Publications, 2003.

[400] von Neumann, J. (1961), “Method in the physical sciences,” in: Collected
Works Vol. VI, Theory of Games, Astrophysics, Hydrodynamics and
Meteorology, A. H. Taub, ed., London (UK): Pergamon Press.

[401] Hofstadter, D. R. (1982). “Analogies and Metaphors to Explain Godel’s

Theorem,” Two-Year College Mathematics Journal, 98-114.

[402] De Regt, H. W., and Dieks, D. (2005). “A contextual approach to scien-

tific understanding,” Synthese, 144(1), 137-170.

[403] Bohr, N. (1949) “Discussion with Einstein on Epistemological Problems

in Atomic Physics” in Shlipp, P. A. (Ed.) Albert Einstein, Philosopher-
Scientist. Cambrifge (UK): Cambridge University Press.

[404] Schrödinger, E. (1926). “Über das Verhältnis der Heisenberg–Born–

Jordanschen Quantenmechanik zu der meinem,” Annalen der Physik,
384(8), 734-756.

[405] Heisenberg’s letter to Pauli, 8 June 1926; quoted and translated in Cas-
sidy, D. (1992) Uncertainty: The Life and Science of Werner Heisenberg.
New York (NY): W.H. Freeman and Co. (p.215).

[406] Michael Weiss: “Anschaulichkeit, Abscheulichkeit” (1992) http://math.

ucr.edu/home/baez/photon/anschaulichkeit.htm

[407] Andersson, E., Barnett, S. M., and Aspect, A. (2005). “Joint measure-
ments of spin, operational locality, and uncertainty.” Physical Review A,
72(4), 042104.

[408] Manners, J. (2000) “Quantum physics: an introduction”. Boca Raton

(FL): CRC Press.

260
[409] Ou, Z. Y. J. (2007). Multi-photon quantum interference. Berlin (DE):
Springer (p. 7).

[410] Wigner, E. P. (1963). “The problem of measurement,” American Journal

of Physics, 31(1), 6-15.

[411] Bell, J. S. (2004). Speakable and unspeakable in quantum mechanics:

Collected papers on quantum philosophy. Cambridge (UK): Cambridge
university press.

[412] Pittman, T. B., Jacobs, B. C., and Franson, J. D. (2005). “Heralding

single photons from pulsed parametric down-conversion,” Optics com-
munications, 246(4), 545-550.

[413] Harder, G., Ansari, V., Brecht, B., Dirmeier, T., Marquardt, C., and
Silberhorn, C. (2013). “An optimized photon pair source for quantum
circuits,” Optics express, 21(12), 13975-13985.

[414] Sun, F. W., and Wong, C. W. (2009). “Indistinguishability of indepen-

dent single photons,” Physical Review A, 79(1), 013824.

[415] Bourdieu, P. (1999). “The specificity of the scientific field,” in Biagioli M.

(ed.) The science Studies Reader. New York (NY): Routledge, pp. 31-50

[416] Bourdieu, P. (1984). Distinction: A social critique of the judgement of

taste. Cambridge (MA): Harvard University Press.

[417] Agenzia Liguria Lavoro – Ente strumentale della Regione Liguria, “I

quaderni dell’OML” 14 (2013) http://www.aligurialavoro.it/prod/oml/
omlquad.asp

[418] Lombardo, F. (2007). “Le seconde generazioni di immigrati ed il concetto

di metisságe”, M.S. Thesis in Diplomatic and International Sciences, Uni-
versity of Genova.

[419] Kelvin, W. T. (1904). Baltimore lectures on molecular dynamics and the

wave theory of light. London (UK): C.J. Clay and Sons.

[420] Mandelstam, L., and Tamm, I. (1945). “The uncertainty relation between
energy and time in nonrelativistic quantum mechanics.” J. Phys.(USSR),
9(249), 1.

[421] Amaldi U., (2003) Corso di Fisica, Vol.3. 5th ed., Bologna (IT): Zanichelli.

[422] Levrini, O. and DiSessa A. (2008). “How students learn from multiple
contexts and definitions: Proper time as a coordination class.” Physical
Review Special Topics-Physics Education Research, 4(1), 010107.

261
 Bibliography

[423] Steinberg, A., Kwiat, P., and Chiao, R. (2006). “Quantum optical tests of
the foundations of physics.” In Springer Handbook of Atomic, Molecular,
and Optical Physics. New York (NY): Springer (pp. 1185-1213).

[424] Petrov, G. M. and Giuliani, J. L. (2003) “Inhomogeneous model of an

Ar–Hg direct current column discharge.” Journal of applied physics,
94(1), 62-75.

[425] http://commons.wikimedia.org/wiki/File:Fluorescent lighting

spectrum peaks labelled.gif

[426] ITU-R Recommendation BT.709-5. International Telecommunication

Union, Geneva, Switzerland, 2002.

[427] Elliott, K. H. and Mayhew, C. A. (1998). “The use of commercial CCD

cameras as linear detectors in the physics undergraduate teaching labo-
ratory.” Eur. J. Phys. 19, 107–117.

[428] http://hyperphysics.phy-astr.gsu.edu/hbase/vision/solirrad.html

[429] Onorato, P., Malgieri, M., and De Ambrosis, A. (2015). “Quantitative

analysis of transmittance and photoluminescence using a low cost appa-
ratus.” European Journal of Physics, 37(1), 015301.

262
List of Publications

REFEREED JOURNAL PAPERS

1. Malgieri, M., Onorato, P., and De Ambrosis, A. (2014). “Teaching quan-
tum physics by the sum over paths approach and GeoGebra simulations.”
European Journal of Physics, 35(5), 055024.

2. Malgieri, M., Onorato, P., Da Ambrosis A. (2015) “A sum over paths ap-
proach to one-dimensional time independent quantum systems.” Ameri-
can Journal of Physics (accepted).

3. Onorato, P., Malgieri, M., and De Ambrosis, A. (2015). “Measuring

the hydrogen Balmer series and Rydberg’s constant with a homemade
spectrophotometer.” European Journal of Physics, 36(5), 058001.

4. Malgieri M., Onorato P., Mascheretti P., De Ambrosis A. (2014) “Pre-

service teachers’ approaches to a historical problem in mechanics.” Physics
Education 49(5), 500–511.

5. Onorato, P., Malgieri, M., and De Ambrosis, A. (2014). “Phase transi-

tions in one-dimensional mechanical models of thermodynamics and the
physics of the Hall bar system.” Physics Letters A, 378(5), 590-596.

6. De Ambrosis, A., Malgieri, M., Mascheretti, P., and Onorato, P. (2015).

“Investigating the role of sliding friction in rolling motion: a teaching
sequence based on experiments and simulations.” European Journal of
Physics, 36(3), 035020.

7. Onorato, P., Malgieri, M., Mascheretti, P., and De Ambrosis, A. (2015).

“The surprising rolling spool: librational motion and failure of the pure
rolling condition.” European Journal of Physics, 36(3), 038002.

8. Malgieri, M., Onorato, P., Mascheretti, P., and De Ambrosis, A. (2013).

“Reconstruction of Huygens’ gedanken experiment and measurements
based on video analysis tools.” European Journal of Physics, 34(5), 1145.

263
 List of Publications

9. Onorato, P., Malgieri, M., Mascheretti, P., and De Ambrosis, A. (2014).

“The surprising rolling spool: experiments and theory from mechanics to
phase transitions.” European Journal of Physics, 35(5), 055011.

10. Onorato, P., Malgieri, M., and De Ambrosis, A. (2015). “Quantitative

analysis of transmittance and photoluminescence using a low cost appa-
ratus.” European Journal of Physics, 37(1), 015301.

11. Malgieri, M., Onorato, P., and De Ambrosis, A. (2015). “Insegnare la

fisica quantistica a scuola: Un percorso basato sul metodo dei cammini
di Feynman.” Giornale di Fisica, 56(1), 55–70.

REFEREED CONFERENCE PAPERS

1. Malgieri, M., Onorato, P., and De Ambrosis, A. (2015) “What is light?
From optics to quantum physics through the sum over paths approach.”
Proceedings of the GIREP-MPTL International Conference, Palermo,
7-12 July 2014 ISBN: 978-88-907460-7-9.

2. Onorato, P., Malgieri, M. and De Ambrosis, A. (2015). “Rolling motion:

experiments and simulations focusing on sliding friction forces.” Pro-
ceedings of the GIREP-MPTL International Conference, Palermo, 7-12
July 2014 ISBN 978-88-907460-7-9.

3. Onorato, P., Malgieri, M., De Ambrosis, A. (2014). “Revisiting Historical

experiments with new technologies: tracking Huygens’ footsteps.” Pro-
ceedings of the 18th MPTL Workshop, Madrid 11-13 September 2013;
MPTL’18–Book of Proceedings ISBN 2-914771-90-8 .

4. Malgieri, M., Onorato, P., and De Ambrosis, A. (2014). “Reconstruction

and video analysis of a thought experiment by Christiaan Huygens.” Sci-
ence Education Research For Evidence-based Teaching and Coherence in
Learning (Proceedings of the ESERA 2013 Conference) ISBN 978-9963-
700-77-6

5. Onorato, P., Malgieri, M., and De Ambrosis, A. (2014). “Learning

about magnetic force: experiments versus theoretical explanations.” Sci-
ence Education Research For Evidence-based Teaching and Coherence in
Learning (Proceedings of the ESERA 2013 Conference) ISBN 978-9963-
700-77-6

264

View publication stats