International Journal of Science Education

ISSN: 0950-0693 (Print) 1464-5289 (Online) Journal homepage: www.tandfonline.com/journals/tsed20

Three practices of experimenting in the physics

classroom – impact on students’ content
knowledge acquisition

J. Winkelmann, S. F. C. Wenzel, M. Ullrich, H. Horz & R. Erb

To cite this article: J. Winkelmann, S. F. C. Wenzel, M. Ullrich, H. Horz & R. Erb (29 Jan
2025): Three practices of experimenting in the physics classroom – impact on students’
content knowledge acquisition, International Journal of Science Education, DOI:
10.1080/09500693.2025.2453954

To link to this article: https://doi.org/10.1080/09500693.2025.2453954

© 2025 The Author(s). Published by Informa

UK Limited, trading as Taylor & Francis
Group

Published online: 29 Jan 2025.

Submit your article to this journal

Article views: 1021

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tsed20
INTERNATIONAL JOURNAL OF SCIENCE EDUCATION
https://doi.org/10.1080/09500693.2025.2453954

Three practices of experimenting in the physics classroom –

impact on students’ content knowledge acquisition
a
J. Winkelmann , S. F. C. Wenzelb, M. Ullrich b
, H. Horz b
and R. Erb c

a
Physics Department, University of Education, Schwäbisch Gmünd, Germany; bInstitute of Psychology,
Department of Educational Psychology, Goethe University Frankfurt, Frankfurt, Germany; cInstitute for
Physics Education, Goethe University Frankfurt, Frankfurt, Germany

ABSTRACT ARTICLE HISTORY

Experimenting is essential in teaching physics. Its importance is Received 26 July 2024
reflected in the large proportion of the teaching time devoted Accepted 3 January 2025
to experimenting. Many teachers and students assume that
KEYWORDS
experimenting promotes acquiring content knowledge. In a Experimenting; content
longitudinal experimental study, seventh- and eighth-grade knowledge; teacher beliefs
students (n = 1507) were examined about their content
knowledge in geometrical optics acquired through different
practices of experimenting in the classroom. The study
distinguished between demonstrations conducted by the
teacher and practical work done by the students with different
levels of guidance. Multilevel analyses showed no differences in
students’ acquired content knowledge with regard to the
applied practice of experimenting. It was also found that
teachers’ constructivist beliefs had a small negative impact,
while transmissive beliefs had a small positive impact on
students’ content knowledge. No interaction between teachers’
beliefs and the different practices of experimenting was found.
Although the results dampen hopes for greater knowledge
acquisition through practical work, they confirm earlier results
by the authors and strengthen findings from teaching-learning
research. This emphasises the significance of classroom process
quality, rather than the distinction between demonstrations and
practical work on the level of structural elements within a
lesson. Educational implications for teaching derived from these
findings are discussed.

Introduction
Experimenting is a common teaching method for science classes (Singer et al., 2006).
Tesch (2005) found that about two-thirds of teaching time in German physics classes
are used for experimenting, including preparation and postprocessing. One practice
of experimenting is practical work of students (NRC, 2012). Teachers regard practical

CONTACT J. Winkelmann [email protected] Physics Department, University of Education,

Schwäbisch Gmünd, Germany
© 2025 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License
(http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any
medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way. The terms on which
this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent.
2 J. WINKELMANN ET AL.

work of students often as central to the appeal and effectiveness of science education
(Abrahams & Millar, 2008; Dillon, 2008). Furthermore, many teachers are confident
that their students will gain a deeper understanding through practical work and
a link between theory and practice would be achieved (de Winter & Millar, 2023;
Lederman et al., 2014; Welzel et al., 1998). Students themselves have similar expec­
tations regarding practical work (Behrendt, 1991; Cerini et al., 2003; Winkelmann
& Erb, 2011). Moreover, they identify the effective use of practical work as the
most significant element that characterises a good physics teacher (de Winter &
Airey, 2020).
Despite high expectations associated with practical work, this practice of experi­
menting is complex on various levels. Practically, students must independently
engage with unfamiliar materials from the physics collection. Socially, they often col­
laborate in small groups. Moreover, it can often be observed that students follow a
step-by-step instruction thereby overlooking the investigative nature of the experimen­
tal process.
An alternative practice of experimenting is the demonstration experiment, where
one person - usually the teacher - performs the experiment in front of the learning
group, and the students (physically) passively follow what is happening. This reduces
the previously formulated aspects of complexity. Studies that have investigated the
effect of different practices of experimenting in a controlled manner are rather rare
and come to contradictory findings. Recent literature reviews on the role of experi­
menting in science classes focus on the impact of experimenting on student academic
performance (the focus of most of the reported studies) and any effects on interest,
motivation, and attitudes towards science (Gericke et al., 2022; Halawa et al., 2020;
Millar, 2010; Osborne et al., 2003; Potvin & Hasni, 2014). Therefore, the present
study aims to compare different practices of experimenting in physics classrooms, dis­
tinguishing between demonstrations and practical work concerning students’ content
knowledge acquisition.
Concerning the different role of the teacher in demonstrations compared to practical
work conducted by students and given the essential role of the teacher on students’ learn­
ing and motivation (Dobber et al., 2017; Hattie, 2009) this study also investigates the role
of teacher beliefs. Well-established models of teachers’ professional behaviour address
not only the teacher’s subject-specific and pedagogical knowledge but also his or her
beliefs about teaching (Baumert & Kunter, 2006). The given study focuses on construc­
tivist and transmissive beliefs of the teacher and the influence of these beliefs on students’
content knowledge acquisition in experiment-based physics lessons. Furthermore, the
interaction between teachers’ beliefs about teaching and the practice of experimenting
has been investigated.
This paper presents the results of a field study (n = 1507) in which the impact of three
different practices of experimenting on the acquisition of seventh- and eighth-grade stu­
dents’ content knowledge was compared. The three practices were (A) demonstrations,
(B) practical work with detailed step by step instruction, and (C) practical work with less
instruction. We consider the development of learning through the provision of domain-
specific opportunities to process content information (Seidel & Shavelson, 2007) and aim
to justify the use of practical work on a topic-specific basis (de Winter & Millar, 2023) –
in this study geometrical optics.
 INTERNATIONAL JOURNAL OF SCIENCE EDUCATION 3

Theoretical background
Experimenting in science classes – a conceptual approximation
The term ‘experimenting’ in the context of science teaching is often used, yet there are
divergent ideas regarding the specific activities encompassed by it. Hacking (1983)
defines experimenting in class as an activity which involves an intervention to produce
the phenomenon to be observed or to test a hypothesis (Hacking, 1983 in Millar,
2010, p. 109). Therefore, the pure observation of a natural phenomenon would not
count as experimenting, as there is no planning and execution of an activity in order
to create a phenomenon. In line with this, experimenting can be structured by several
stages, e.g. planning, execution and evaluation (Klahr et al., 2011; Schreiber et al.,
2009). In school contexts, each of these stages can be opened up for more or less inde­
pendent activities by students.
Teaching concepts for active learning in science classrooms were mainly proposed in
inquiry-based learning (NRC, 1996). Inquiry-based teaching includes students drawing
upon their scientific knowledge to ask scientifically oriented questions, collect and
analyse evidence from scientific investigations, develop explanations of scientific
phenomena, and communicate those explanations with their teacher and peers (NRC,
1996). However, the term inquiry is controversial, as there are different ideas about its
meaning and ambiguities regarding the implementation (Crawford, 2014; Furtak et al.,
2012a; Hmelo-Silver et al., 2007). Therefore, the description of concrete science practices
– such as conducting experiments – is assumed to enable better communication of
meaning amongst professional science educators (NGSS, 2013; NRC, 2012; Osborne,
2014).
In experimenting, following Gericke et al. (2022) and Banchi and Bell (2008), three
different instructional practices can be distinguished about teacher regulation (struc­
tured/guided/open inquiry). Gericke et al. (2022), show, that more open practices of
experimenting are not always superior to more teacher regulated ones, as indicated by
lots of previous studies (Caglak, 2017; Dobber et al., 2017; Furtak et al., 2012; Heindl,
2019). Moreover, the teacher’s support appears to be essential for students before,
during, and after an exercise, as implied in previous reviews (Agustian & Seery, 2017;
Akuma & Callaghan, 2019; Lunetta et al., 2007).
Models of teaching and learning characterise learning as a self-regulated and construc­
tive process (Bransford et al., 2000; Collins et al., 1989). In science education research, a
constructivist view of learning has become established. Accordingly, experimenting
should enable (cognitive) activation of learners (Barron & Darling-Hammond, 2010;
Hood Cattaneo, 2017). Hence, this paper focuses on a constructivist perspective as the
compared experimenting situations differ in the student’s actions during planning and
executing the experiments. In line with the findings of Gericke et al. (2022), research
from educational psychology and physics education (Hofstein & Lunetta, 2003; Mayer,
2004) suggests that the problems that learners might face when experimenting can be
solved more easily when learners are provided with a guiding instruction. Instructions
should offer steps that are not too detailed, as instructions involving too detailed gui­
dance are not effective (Kirschner et al., 2006; Schmidt et al., 2007). On the other
hand, De Jong and van Joolingen (1998) report that students encounter problems
when they follow too vague instructions. The instruction seems most effective if it
4 J. WINKELMANN ET AL.

offers a mixture of autonomous research – no instruction at all – and detailed instruction

to solve the problem. However, Hodson and Bencze (1998) fear that students who follow
a recipe instruction develop a false picture of science. In addition, it has been shown that
students do not think deeply about the meaning of their acts when they follow close
instructions (Hart et al., 2000).

Effects of experimenting on learning

Science education research has grown interested in independent experimenting since the
1960s (Trumper, 2003). Rönnebeck et al. (2016) emphasise that studies vary in their focus
on two essential dimensions of inquiry: the types of activities students engage in and the
degree of guidance teachers provide. For example, Dekkers and Thijs (1998) report a
higher efficiency in learning through demonstration small group experimenting.
Jerrim et al. (2019) report that ‘high levels of inquiry or unguided inquiry have no relation­
ship with attainment at all’ (p, 42). Also, it was shown in several reviews that practical
work in labs can only make small contributions to an improved understanding of
natural sciences (Hofstein & Lunetta, 1982; Shulman & Tamir, 1973; Singer et al.,
2006). Another review with a sharper focus on developing content knowledge reports
no clear evidence that practical work offers an advantage over demonstrations in knowl­
edge acquisition (Millar, 2010). In a recent review of 12 international studies, Halawa
et al. (2020) find that teaching science by guided inquiry has positive effects on students’
science learning concerning both conceptual knowledge and procedural skills, which
stands in line with former reviews (Furtak et al., 2012; Heindl, 2019; Lazonder &
Harmsen, 2016). However, Dobber et al. (2017) claim that teachers and students need
distinct training to adapt to the demands required for inquiry teaching. Small to
medium effect sizes for and against practical work can be found in the literature.
Studies that focus on the effects of experimenting on learning in science classes are chal­
lenging to compare. Studies often use different designations of the same concepts (inquiry
learning, laboratory work, practical work, hands-on situations, experimental situations,
…). Many studies focus on different subject areas, pursue very different research
approaches (qualitative or quantitative) and look at different dependent and independent
variables. Table 1 does not claim to be a systematic literature review, but impressively shows
the wide methodological range of existing studies. In addition, Hofstein and Lunetta (2003)
have identified partially insufficient designs with regard to the control of procedures, the
reporting of teaching actions and the sample size, especially in quantitative studies.
Potvin and Hasni (2014) emphasise that only a few studies have ‘practical work’ as a
pure subject of investigation. Schwichow et al. (2016) deduce that experimenting alone
cannot be decisive for learning success, but that special attention should be paid to the
task instead of the material.
Other studies in which practical work or demonstration experiments were directly
compared (Hopf, 2007) suggest that differences in the organisation of lessons, i.e. the
structural, surface-level characteristics of a lesson, are not decisive, at least in terms of
efforts to enhance content knowledge acquisition. Instead, what is more significant are
depth characteristics and the teachers themselves (Hattie, 2009; see also the discourse
of the terms ‘hands-on’ and ‘minds-on’ in Furtak & Penuel, 2019 and the response by
Osborne, 2019).
Table 1. Overview of the wide methodological range of studies regarding the impact of experimenting practices.
Research Age of
Study Content Domain approach Dependent variable Independent variable Treatment lable sample
Abrahams & science various qualitative effectiveness of „typcial“ science lessons practical work secondary
Millar, 2008 practical work schools
Berg et al., chemistry Comparing catalysts qualitative & learning outcomes open-inquiry vs. an expository open-inquiry universtiy
2003 quantitative depending on version of a laboratory level
attitudes experiment
Högström chemistry modern materials qualitative learning in laboratory student-teacher interactions laboratory work grade 8
et al., 2010 contexts
Shana & chemistry & biology photosynthese, endothermic quantitatiev academic attainment traditional teaching method vs. practical work grade 10–11
Abulibdeh, and exothermic reactions, in science intensive practical work
2020 Acid-Base Titration
Hopf, 2007 physics electricity and optics quantitative content knowlegde Design of student experiments Student grade 7–8
and interest (openness, problem orientation, experiments
authenticity)
Maričić et al., integrated science air-related content quantitative content knowledge teacher demonstration vs. hands- hands-on grade 3
2019 (biology, physics, on experiments vs. verbal- experiments
chemistry, textual method
ecology)
Rumann, 2005 chemistry acidic and alkaline solutions quantitative content knowledge cooperative work in small groups student Grade 7
vs. teacher-centered classroom experiment
discussion
Blanchard science forensic quantitative knowledge of content, guided inquiry – based instruction guided inquiry – middle and
et al., 2010 procedure and vs. traditional, laboratory based high school
nature of science instruction instruction
Schulz, 2011 Chemistry alcohols quantitative content knowledge, quality features of experiments in grade 10
scientific methods, experimentation phases chemistry
interest classroom
Potvin et al., science and various quantitative interest teaching segments taught „as inquiry based grade 5–10
2017 technology usual“ vs. segments enhanced interventions
with inquiry-based
characteristics
INTERNATIONAL JOURNAL OF SCIENCE EDUCATION
5
6 J. WINKELMANN ET AL.

Role of teacher beliefs

Numerous studies show that the influence of teacher’s characteristics (e.g. beliefs, peda­
gogical-content knowledge, instructional practices) is decisive for the success of the
lesson and exceeds the influence of other aspects (Högström et al., 2010; Staub &
Stern, 2002; Velasco et al., 2016; Wallace, 2014). The essential role of the teacher is docu­
mented concisely by meta-studies (Dobber et al., 2017; Hattie, 2009). In addition, there is
some evidence for an interaction between teacher characteristics and experimental prac­
tice as the findings of Winkelmann and Erb (2018) and Winkelmann (2015) suggest.
Despite the substantial research on teacher beliefs and their (at least assumed) impact
on students’ learning and motivation, a precise definition is still missing that most
researchers can agree upon (Skott, 2015). Concerning the objects or phenomena, tea­
chers’ beliefs apply to Baumert and Kunter (2006) suggested a pragmatic approach.
They identified epistemological beliefs and subjective theories about teaching and learn­
ing as crucial for general instructional quality. According to Voss et al. (2013), epistemo­
logical beliefs and beliefs about teaching and learning can be attributed to a more generic
belief system that complies with constructivist or transmissive theories. Oettinghaus et al.
(2012) and Lamprecht (2011) suggest that constructivist beliefs are strongly associated
with students’ practical work in science teaching, while teacher demonstration exper­
iments are more in line with transmissive beliefs. Although research on teacher beliefs
has increased substantially in the past 25 years, the relationship between students’ learn­
ing outcomes, teaching practices, and teacher beliefs remains to be determined. Seidel
et al. (2008) found no relationship between teachers’ constructivist or transmissive
beliefs and their teaching actions. They also failed to demonstrate any influence of
teacher beliefs on students’ learning outcomes. In addition, Kunter et al. (2013) found
that constructivist beliefs and classroom management skills correlated negatively since
other studies indicate direct effects of teachers’ epistemological beliefs (Staub & Stern,
2002; Stipek et al., 2001; Voss et al., 2013; Yang et al., 2020).

Research questions
The current research on science education presents a conflicting picture regarding the
impact of experimenting practices. While some studies and reviews (e.g. Halawa et al.,
2020; Lazonder & Harmsen, 2016) report positive effects of guided, inquiry-based learn­
ing on the understanding of scientific concepts and skills, other studies (e.g. Millar, 2010;
Singer et al., 2006) show no clear advantage of practical work over demonstrations. These
contradictory findings could indicate that the effectiveness of experiments is highly
dependent on context, lesson design, and setting. Research is also complex to compare
as many studies exhibit a high degree of variability in their methodological approaches
(see Table 1). At the same time, however, teachers and students have high expectations,
especially for practical work.
The present study builds on these findings with a controlled experimental design and a
focus on beginning physics classroom in an optics context. To answer the following
research questions, an experimental field study with a large sample was conducted. On
the one hand, this study should provide results transferable to real school lessons. On
the other hand, the treatments were to be compared with the largest possible control
 INTERNATIONAL JOURNAL OF SCIENCE EDUCATION 7

of confounding variables and have been strongly pre-structured (see the following
section).
RQ.1 To what extent do students in grade 7 and 8 differ in their physics content knowledge
acquisition (optics) when comparing three different practices of experimenting?

The enormous importance of the teacher for learning has been discussed above. There­
fore, the participating teachers’ beliefs about teaching and learning physics were measured
to consider their influence. This aims to address the second and third research questions.
RQ.2 To what extent do the teachers’ beliefs about teaching and learning physics influence
the students’ content knowledge acquisition when comparing three different practices of
experimenting?

RQ.3 To what extent is there interaction between teacher beliefs about teaching and learning
physics and the performance of different experimental practices concerning students’
content knowledge acquisition?

Methods and design

A field study with a pre-posttest design was conducted to ensure the results’ applicability
in real classroom settings. Students stayed in their original classes, taught by their regular
teachers during school hours. Methodologically, control measures were implemented (on
student level: gender, age, prior knowledge; on teacher level: age, professional experience;
on a third level: school type) including standardisation of materials, random allocation of
classes to experimenting practices, and ensuring comparability through teacher training,
scripted lessons, and uniform worksheets for students.

Intervention in physics classrooms

The study was carried out as part of regular physics lessons at 7th and 8th-grade high
schools, comprising three 90-minute sessions per week (2 lessons of 45 min) on light
refraction. Optics was chosen as a subject due to its easily visualised phenomena, facil­
itating comparability across practices of experimenting. Lessons began with a 5-
minute introduction, followed by a 30-minute experimental phase, and ended with a
10-minute class discussion to reinforce observations. Several different experiments
about light refraction were worked on, omitting routine tasks and repetition phases
for consistent conditions. Table 2 summarises the course of the lessons.
The experiments aimed to construct new physics knowledge, though they were stan­
dard in the field of light refraction. All of the experiments were real experiments

Table 2. Overview of the experiments used. Every lesson lasts 90 min.

Lesson Topic Experiment
1st lesson Exploring light refraction 1. Aiming-experiment: Catching the ‘fish’
2. Aiming-experiment: Pins in a row
2nd lesson Law of light refraction and total reflection 3. Light refraction with table optics
4. Total reflection with table optics
3rd lesson Light through a converging lens 5. Focusing light in the ‘Cobbler’s-bowl’
Dependency of light focusing 6. Light refraction with converging lens
8 J. WINKELMANN ET AL.

conducted by the teacher as a demonstration or in small groups by the students (usually

three students in one group). Evaluation of experimental data by the teacher was not con­
sidered experimenting, nor was independent drawing of conclusions from observations.
Following the considerations in the theoretical background, three different practices of
experimenting were selected and are described in more detail below. All practices focused
on experiments suitable for both demonstrations and practical work, aiming for
maximum design similarity (similar material, worksheets and duration).
The present study compares three practices of experimenting with varying student
activity, focusing on the impact on students’ content knowledge acquisition in physics.
While ensuring comparability, the main distinction of experimenting practices lies in
the transmissive nature of demonstrations versus the constructivist approach of practical
work. The two practical work situations differ in terms of scaffolding in the development
of the experiment.
The different explored practices of experimenting were:

(A) Demonstration: The teacher conducts an experiment in front of the entire class.
Students note their observations for every step of the experiment on a worksheet
but remain passive otherwise. The learning of the respective physical content is
understood here as a transmissive transfer of knowledge by the teacher.
(B) Practical work with detailed instruction: Students were guided step by step
through their practical work, provided with a problem situation was given as well
as the experimenting materials.
(C) Practical work with less instruction: Students received relatively open instruction,
encouraging autonomous thinking. The instruction was not minimal but less elab­
orate than in group B. Students were given a problem situation as well as the experi­
menting materials and had to plan and carry out the experiment independently.
Teachers were allowed to assist if there was a risk of students failing the experiment.

Teachers reported only one situation in which students deviated from the expected
planning and execution of the experiments (Exp. 1, see Table 2). This did not lead to
a failure of the experiment but could be brought up in the class discussion during the
evaluation phase: Refraction of light can only be observed at an angle of incidence
greater than 0°.
Following Gericke et al. (2022), treatment A and B correspond to ‘structured inquiry’,
while treatment C roughly aligns with ‘guided inquiry’. The three practices (A-C, summar­
ised in Figure 1) serve as the independent variables in the study, derived conceptually on
the literature described above and consistent with a previous study of Winkelmann (2015).
To illustrate the experiments conducted, a detailed example from the first lesson, ‘pins
in a row,’ is presented in Figures 2 and 3. Following this initial lesson, further exper­
iments provide a more detailed physical investigation and description of light refraction.
Students measure refraction angles and learn about the law of refraction.

Data collection and sample

The study was conducted in the federal states of Hessen and North Rhine-Westphalia,
which both offer a typical German secondary school system with three tracks
 INTERNATIONAL JOURNAL OF SCIENCE EDUCATION 9

Figure 1. The three practices of experimenting differ in the students’ activity during the stages plan­
ning and execution of the experiment.

(comprehensive school = low and middle track; Gymnasium = high track). Since there
was interest in the analysis of cross-level interactions, the number of groups (classes)
had to be high enough to gain the required power. Hox (2002) suggests about 50
classes with about 20 individuals per class as a rule of thumb to analyse cross-level inter­
actions. A sample size of 60 classes (20 per practice of experimenting) appeared to be
sufficient to obtain an equal distribution on three different practices of experimenting.

Figure 2. a-b. An experiment to get to know the phenomenon of light refraction; practical work.
10 J. WINKELMANN ET AL.

Figure 3. a-b. An experiment to get to know the phenomenon of light refraction; demonstration.

The data collection took place in 2018. The participating teachers received materials for
planning, conducting, and documenting the experiments, and paper-and-pencil test
booklets for conducting the evaluation survey via mail. Stamped packages were provided
for the return of all materials. Finally, we recruited 62 classes from 28 schools with 1507
students (N = 757, female; N = 704, male; N = 46, no statement) between 12 and 16 years
old (M = 13.7, SD = 0.8). As the time of teaching geometrical optics varies across schools,
70,8% of the students (N = 1067) were in seventh grade, while 29.2% (N = 440) were in
eighth grade. Classes from high schools and comprehensive schools took part (N = 46,
high school classes; N = 16, comprehensive school classes), with most participating
schools being high schools (N = 23, high schools; N = 5, comprehensive schools). On
average, a class consisted of 24 students (min = 14, max = 32). The allocation of the
classes to the experimenting practices is roughly the same (N = 18, A – Demonstration;
N = 22, B – practical work with detailed instruction; N = 22, C – practical work with less
instruction). Of the 62 participating teachers, 38.7% were female, and 58.1% were male
(3.2%, N = 2, no statement). The teachers were between 28 and 65 years old (M = 46.3,
SD = 11.4) and had between zero and 39 years of professional experience (M = 17.3,
SD = 11.5).

Instruments and variable descriptions

Students’ test-Scores on content knowledge
To measure students’ content knowledge, an Item-Response-Theory (IRT)-scaled test
(German language) on geometrical optics was used (Weber et al., 2018) that in parts
resorted to items from Winkelmann (2015). The test consisted of multiple-choice ques­
tions and provided partly different subsets of items for the different points of measure­
ment prior to the lectures on geometrical optics (pre-test with 20 items) and afterward
 INTERNATIONAL JOURNAL OF SCIENCE EDUCATION 11

(post-test with 20 items, including 4 pre-test items as anchor items). An example item
was: ‘A stick is held in a container filled with water. If you look at it from above, the
stick seems kinked. How do you see the same stick when you look at the container from
the side?’. By using IRT-modelling, scores from partly different subsets of items drawn
from a larger item pool can be mapped on the same scale and set about each other
(Cella et al., 2007).
IRT-estimates of students’ content knowledge were obtained based on test responses
and a latent regression model that integrates several covariates (von Davier & Sinharay,
2014). The expected-a-posteriori/plausible-values (EAP/PV) reliability of the IRT-esti­
mates of students’ content knowledge prior to the treatment (CKprior) is .60 with a
mean of item difficulties of 0.15. The mean of item difficulties of the test applied after
the lectures on geometrical optics is 0.80, which implies that the test consisted of some­
what more complex items. The reliability of the IRT-estimates of students’ content
knowledge after the lectures (CK) is .54.

Teachers’ beliefs
Therefore, prior to the beginning of the geometric optics lessons observed in this study,
teachers’ beliefs about teaching and learning physics were assessed using two belief scales,
addressing ‘Constructivist understanding of teaching and learning’ (CB, ten items) and
‘Transmissive understanding of teaching and learning’ (TB, seven items) (Lamprecht,
2011; Oettinghaus et al., 2012). An example item for CB is: ‘A good physics class
usually involves students working experimentally.’. An example item for TB is: ‘Teachers
should teach detailed procedures for conducting experiments.’. CB and TB were measured
on a 5-point likert-scale with response options from 1 (I strongly disagree) to 5 (I strongly
agree). For reliabilities and descriptive statistics, see Table 4.

Control variables
As the study was taking place in schools to enable the students to learn in their familiar
environment, control variables on the students’ level, the class e.g. teachers’ level and the
school level were considered.
Student Level. The students’ test score on content knowledge assessed prior to the
beginning of the geometric optics lessons observed in this study was used as a key
control variable for content knowledge on geometrical optics after the lectures. Addition­
ally, the sociodemographic characteristics ‘gender’ and ‘age’ of the students were
included as control variables in the analyses.
Class level. At the class level, the teachers’ gender, as well as teachers’ age and work
experience, and class size were used as control variables.
School level. At the school level, the type of school (high school or comprehensive
school) was included in the analyses as a control variable.

Statistical analysis
Treatment of missing values
The collected data set contains missing values. The percentage of missing data in the
control variables at the student level is 3.05% for gender and 5.97% for age. At teachers’
level, there are 2.99% missing values for gender, 13.60% for age, and 14.53% for work
12 J. WINKELMANN ET AL.

experience. On the scales regarding teachers’ beliefs 14.53% of answers are missing.
There are no missing values in the students’ test scores on content knowledge (CK
and CKprior) nor the control variables class size and school type. Instead of deleting
cases with missing values (list-wise deletion), which reduces sample size and statistical
power (Peugh & Enders, 2004), multiple imputation is used to estimate the missing
values. Based on the assumption of missing at random (MAR; Lüdtke et al., 2007), the
state-of-the-art multiple imputation procedure was chosen to handle missing values to
improve the statistical precision of the analyses (see Newman, 2014). To impute dichot­
omous and categorical variables, logistic methods (logistic and polytomous logistic
regression) were used. Missing values on metric variables were imputed using the predic­
tive mean matching (pmm) approach. Twenty datasets were generated in which missing
values were replaced by estimated values, conducting 15 iterations per imputation (van
Buuren, 2018).

Multilevel analysis
The aim of the present study was to investigate the effects of (RQ.1) different experimen­
tal situations in physics classrooms, (RQ.2) teacher beliefs about teaching and learning
physics, and (RQ.3) a possible interaction between experimental situations and teacher
beliefs, on students’ acquired content knowledge. Multilevel analyses were the method
of choice to account for the hierarchical structure of the present data set. Therefore,
several hierarchical linear regression models were estimated. The models were built on
each other and adjusted for the effects of prior knowledge.
First, in an unconditional null model, interclass correlation coefficients (ICCs) were
estimated, representing the proportion of variance in latent school and class effects by
indicating the variance in the dependent variable attributable to the difference between
schools and classes (Hox et al., 2017). Analysis of the unconditional model suggests
low variation in students’ test scores on content knowledge between classes (ICC
for classes: .013) and slightly higher variation between schools (ICC for schools:
.020). Even though the ICCs are pretty low, we still decided to conduct 3-level
models to investigate the research questions (similar values also reported in Hedges
& Hedberg, 2013).
With Model 1, the effect of treatment on students’ CK, adjusted for CKPrior, was
analysed. Model 2 additionally included teachers’ constructivist and transmissive
beliefs, which might explain variations in students’ content knowledge. With Model
3, a potential interaction between teachers’ beliefs and the treatment was analysed.
As Model 3 did not significantly fit the data better than Model 2, a fourth model was
estimated. In this final step, Model 4 equals Model 2 but includes all previously
described control variables at student, class, and school levels. The variables represent­
ing teachers’ understanding of teaching and learning (constructivist = CB and trans­
missive = TB) and CKPrior were centred before being used in the analyses. CKPrior,
as a level 1 variable, was group mean centred, whereas CB and TB on level 2 were
grand mean centred. With R software version 4.1.1. (R Core Team, 2021) the
package ‘lme4’ (Bates et al., 2015) was used to conduct all multilevel analyses. For
model comparisons, the package ‘mitml’ version 0.4-3 (Grund et al., 2021) was used
to calculate the Wald-test directly for each data set and then pool the resulting Chi²
values (Li et al., 1991).
 INTERNATIONAL JOURNAL OF SCIENCE EDUCATION 13

Table 3. Level 1: Mean, standard deviation and correlations of the student level variables.
Correlations
Scales/Variables Mobs SDobs Mimp SDimp 1 2 3 4
1 Content knowledge (CK) 0.02 0.36 0.02 0.36 – .83*** −.10*** −.10***
2 Content knowledge prior to −0.02 0.35 −0.02 0.35 .83*** – −.05 −.04
treatment (CKprior)
3 Students age 13.70 0.83 13.71 0.83 −.10*** −.04 - −.03
4 Students gender (0 = m, 1 = f) 0.52 - 0.52 - −.10*** −.04 −.03 -
Notes. obs = Characteristic values calculated with the observed incomplete data. imp = Pooled parameter estimates
obtained from imputed data. Below the diagonal, the correlations between the variables in the observed incomplete
data set with pairwise case exclusion are plotted. Above the diagonal, the correlations between the variables in the
pooled data are displayed across the imputed data sets.
*** p < .001 (two-sided).

Results
Tables 3 and 4 show the mean values (M), standard deviations (SD), reliability coeffi­
cients (McDonald’s omega), and correlations of the variables on student and teacher
levels used in the multilevel analysis. Both the characteristic values calculated based on
the observed incomplete data and the imputed data are presented.
Table 5 shows the results from the multilevel modelling analyses. The null model
shows that the largest variation in students’ CK is due to differences between students.
A variation in students’ CK between classes and schools nevertheless exists. 1.3% of
the variation in students’ CK can be ascribed to differences between classes and 2.0%
to differences between schools, while the rest of the variation is due to individual charac­
teristics of the students.
Model 1 indicates that students’ CK prior to the treatment is significantly related to
students’ CK after the treatment. Students with higher CK before the lecture on geo­
metrical optics also show higher CK scores after the lecture. The form of the experiment­
ing situations experienced in class has no impact on students’ CK. We enhanced the
model in three steps. Finally, Model 4 shows that prior CK still has the biggest effect
on students’ CK. In addition, teachers’ CB still has a small negative and teachers’ TB
has a small positive impact on students’ CK, both now statistically significant. Further­
more, students’ gender has a small negative impact on students’ CK, meaning that
girls have slightly lower scores on CK than boys. School type is also a significant
control variable, indicating that students from comprehensive schools have lower
scores on CK than students from high schools. The other control variables incorporated
in Model 4 did not correlate significantly with students’ CK. Regarding model fit, Model
4 is significantly better than Model 2, indicating that including control variables at indi­
vidual, class, and school levels proves relevant. Thus, the final Model 4 revealed the best
model fit.

Discussion
Summary and interpretation of the results
The present study investigated how different experimental situations in physics classes
affect students’ content knowledge acquisition. The study was conducted according to
a comprehensible design with a high degree of control of confounding variables and
 14
J. WINKELMANN ET AL.

Table 4. Level 2: Mean, standard deviation, reliability coefficients and correlations of the teacher variables.
Correlations
Scales/Variables Mobs SDobs Relobs Mimp SDimp Relimp 1 2 3 4 5
1 Constructivist beliefs of teaching and learning (CB) 4.11 0.53 .90 .85 – −.07 −.03 .06 .27*
2 Transmissive beliefs of teaching and learning (TB) 2.93 0.66 .89 .82 −.07 - .34** .32** .15
3 Teachers age 46.11 11.23 - 46.11 11.30 - −.01 .31* - .85*** −.02
4 Teachers work experience 17.18 11.30 - 16.81 11.33 - .06 .26* .85*** - .15
5 Teachers gender (0 = m, 1 = f) 0.39 - - 0.39 - - .28* .10 −.01 .15 -
Notes. obs = Characteristic values calculated with the observed incomplete data. imp = Pooled parameter estimates obtained from imputed data. McDonald’s omega was calculated as reliability
coefficient. Below the diagonal, the correlations between the variables in the observed incomplete data set with pairwise case exclusion are plotted. Above the diagonal, the correlations
between the variables in the pooled data are displayed across the imputed data sets.
* p < .05 (two-sided). *** p < .001 (two-sided).
 INTERNATIONAL JOURNAL OF SCIENCE EDUCATION 15

Table 5. Results from the multilevel modelling analyses: effect on students CK.
Model
Model 3
Model 1 Model 2 + interaction of Model 4
Null Treatment and + Teacher Treatment and Teacher Model 2 + Level 1 and
Level and variable model prior CK Beliefs Beliefs Level 2 covariates
Level 1 (Students)
Intercept .045 .043 (.048) .039 (.045) .035 (.046) .248 (.196)
(.033)
Students’ CKPrior .930 (.016) *** .931 (.016) .931 (.016) *** .925 (.016) ***
***
Level 1 control
variables
Gender −.036 (.009) ***
Age −.007 (.008)
Level 2 (Teachers/
Classes)
Treatment2 −.008 (.053) .003 (.047) .011 (.050) −.011 (.046)
Treatment3 .013 (.052) .020 (.046) .022 (.048) .008 (.045)
Teacher Beliefs (CB) −.074 −.059 (.030) −.069 (.026) *
(.026) *
Teacher Beliefs (TB) .039 (.020) .036 (.039) .044 (.021) *
Level 2 interactions
Treatment2*CB −.068 (.068)
Treatment3*CB −.019 (.075)
Treatment2*TB .003 (.050)
Treatment3*TB .003 (.054)
CB*TB .012 (.047)
Level 2 control
variables
Gender −.005 (.036)
Age −.001 (.002)
Work experience −.002 (.002)
Class size .001 (.005)
Level 3 (Schools)
School type −.249 (.063) ***
Additional information
ICC Level 2 .013
ICC Level 3 .020
Number of 1 4 6 11 13
estimated parameter
Wald-test pooled 1,184 *** (M0) 6.375 ** 0.227 (M2) 4.614 *** (M2)
Chi² (M1)
Notes. Wald-tests were performed to compare the respective model with the last calculated model, indicated in parenth­
eses.
* p < .05. ** p < .01. *** p < .001.

took place simultaneously in actual physics classes. In this way, the results should be as
usable as possible in practice. By considering the nested structure of the data at three
levels (Table 5), we can see that Students’ prior content knowledge was the strongest pre­
dictor of their content knowledge after the lessons on geometrical optics. There were no
differences in the students’ content knowledge acquisition, due to the different experi­
menting practices [RQ 1]. These findings indicate that demonstration experiments
enable students to acquire knowledge to the same extent as practical work with a high
or low degree of instruction. This confirms the results of a previous study (Winkelmann,
2015; Winkelmann & Erb, 2018). In our study we focused on the first two stages of
experimenting (‘planning’ and ‘execution of the experiments’). During planning and
execution, the students’ actions were structured by worksheets, and the students were
16 J. WINKELMANN ET AL.

repeatedly asked to take notes independently - which, according to the literature, may be
helpful (Gericke et al., 2022). Gericke et al. (2022) also point out that guided inquiry
would be the first choice in experimenting. About the development of content knowl­
edge, we would instead advocate greater serenity in the choice of experimental practices.
In other words, although the learning environment in practical work is more cognitively
demanding (own action, unfamiliar objects, reading instructions, setting up the exper­
iment), the students in this treatment group do not perform worse: this reinforces the
importance of practical work. In addition, the age of the learners could also play an
important role in the practice of experimenting. As we focussed students in grade 7
and 8, it is possible that the potential of practical work (in the sense of constructivism,
i.e. the independent development and networking of knowledge) can only unfold at an
older age.
The influence of teachers’ beliefs about teaching and learning physics was also
included in the analyses. According to the standardised regression-coefficients of
model 4 constructivist teachers’ beliefs have a very small negative and transmissive tea­
chers’ beliefs have a very small positive effect on students’ content knowledge [RQ 2].
This contradicts the results of the video study by Seidel et al. (2008) but fits with the
findings by Baumert and Kunter (2006). The influence of teachers’ beliefs on students’
content knowledge remains even with the addition of control variables at the student,
teacher/class, and school level. Prior knowledge is still the strongest predictor of
content knowledge.
As far as the authors can tell from the available studies, the present study is the first
time that a teacher-side variable (beliefs) has been linked to practical teaching
methods (practical work vs. demonstration) and empirically investigated. We see a poss­
ible explanation for the reported findings in that the students in our sample explicitly
experienced physics as a subject for the first time in their school careers. The small posi­
tive influence of transmissive teachers’ beliefs on the students’ content knowledge could
thus be interpreted as a stronger orientation and guidance, preferred by such teachers
and needed by the students. We studied students in grades 7 and 8, and it seems to us
an interesting further line of research is whether constructivist beliefs of teachers are
helpful only for older learners.
A possible explanation for the low negative influence of constructivist beliefs in our
study could be as follows. Teachers believe a constructivist learning environment
alone is sufficient for students to learn successfully. It is possible that teachers with
high constructivist beliefs intervene too infrequently or support too seldom to turn
active doing into active thinking during practical work, for instance.
Although the effects of teachers’ beliefs are small, they indicate the teachers’ impor­
tance in the classroom. An interaction between beliefs about the teaching and learning
of physics and the practice of experimenting used in the classroom was reported in
our preliminary work (Winkelmann, 2015). However, it could not be explained
then, as the focus was not on teacher characteristics. In the present study, there
would have been an opportunity to explain any confirmation of the results at that
time with the help of teacher beliefs. However, an effect of the interaction of
teacher beliefs and different experimenting practices on the students’ CK could not
be found [RQ 3].
 INTERNATIONAL JOURNAL OF SCIENCE EDUCATION 17

Limitations and further directions

It should be noted that both the studies reported by Gericke et al. (2022) and our results
were collected at class level. Individual performance outcomes in heterogeneous learning
groups have yet to be sufficiently investigated. For the approach to be successful, adher­
ence to the strategy seems critical; if the guided inquiry is not carried out as intended, it
may be less effective than the confirmatory or structured inquiry. In our study, teachers
were asked to retrospectively fill out a questionnaire after the lesson (during the student
posttest) regarding the implementation of the study. No indications of deviations from
the planned lesson plan were to be noted. On this basis, teachers rather closely followed
the script.
Regarding the overall small effects of teacher beliefs, it should be noted that beliefs are
not a bi-polar construct. Teachers may hold more constructivist beliefs in one situation
and more transmissive beliefs in another.
Due to the high expectations that students and teachers have of practical work, the
present study focused on conducting the experiments. The variation of the acting
person shows no influence on the students’ learning. Presumably, interpreting the col­
lected data or the observations contributes to learning to a higher degree than mere
doing. However, we deliberately kept this phase of the experiment the same. However,
it should also be emphasised that the lessons were strongly pre-structured. This
appears justified from a research perspective because it focuses on practices of experi­
menting and ensures the best possible comparability of the treatments. Actual lessons
will undoubtedly be more responsive to student comments and, therefore, less linear.
Also, the lessons in our study did not include any phases of practising and consolidating
what had been learned.

Implications for teaching

Demonstrations are usually carried out by teachers - i.e. experts - so that it can be
assumed that what the students are supposed to observe is shown and observable. In
other words, the experiment ‘works.’ In addition, students can concentrate on observing
and do not have to arrange mostly unfamiliar objects by themselves meaningfully. On the
other hand, this practice of experimenting requires a high level of attention from the stu­
dents. In practical work, the students work independently in small groups. Through
intensive exposure to experimental materials, learning of a scientific phenomenon can
be experienced directly. Observations are linked to situations the students encounter
and are thus better remembered. This corresponds strongly to the idea of constructivist
learning. Our results indicate that it does not matter for students in grades 7 and 8 how
experiments are conducted. Both demonstrations and practical work were justified prac­
tices in science education to promote content knowledge. Since we could not find any
interaction between teacher beliefs and the practice of experimenting [RQ 3], the
freedom of choice for teachers appears to be strengthened. It seems harmless to use
experimentation methods that contradict one’s beliefs. Each teacher should choose the
type of experiment they prefer, but a variety of practices should be performed. Thus,
our study supports the conviction that it is more important how an experiment is inte­
grated into lesson planning than according to which practice an experiment is conducted.
18 J. WINKELMANN ET AL.

Conclusion
Experimenting is a crucial practice in science classes. Contrary to high expectations of
practical work, no advantage (but also no disadvantage) to experimental demonstrations
could be found in this extensive field study about content knowledge development. With
regard to conducting experiments in class, we advocate a calm decision-making process.
There is no one right form of experimenting. Various experimenting practices may be
more purposeful than believing in the beneficial power of practical work.

Acknowledgment
We would like to thank our colleague Jeremias Weber, who played a key role in recruiting the par­
ticipating teachers and their classes, as well as in collecting the data.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was supported by German Federal Ministry of Education and Research (BMBF).

ORCID
J. Winkelmann http://orcid.org/0000-0002-2207-7987
M. Ullrich http://orcid.org/0000-0003-4484-1023
H. Horz http://orcid.org/0000-0002-5173-0252
R. Erb http://orcid.org/0009-0000-6538-9529

References
Abrahams, I., & Millar, R. (2008). Does practical work really work? A study of the effectiveness of
practical work as a teaching and learning method in school science. International Journal of
Social Economics, 30(14), 1945–1969. https://doi.org/10.1080/09500690701749305
Agustian, H., & Seery, M. (2017). Reasserting the role of pre-laboratory activities in university
chemistry laboratories: A proposed framework for their design. Chemistry Education
Research and Practice, 18(4), 518–532. https://doi.org/10.1039/C7RP00140A
Akuma, F. V., & Callaghan, R. (2019). A systematic review characterizing and clarifying intrinsic
teaching challenges linked to inquiry-based practical work. Journal of Research in Science
Teaching, 56(5), 619–648. https://doi.org/10.1002/tea.21516
Banchi, H., & Bell, R. (2008). The many levels of inquiry. Science and Children, 46(2), 26–29.
https://www.jstor.org/stable/43174976
Barron, B., & Darling-Hammond, L. (2010). Prospects and challenges for inquiry-based
approaches to learning. In H. Dumont, D. Istance, & F. Benavides (Eds.), The nature of learning.
Using research to inspire practice (pp. 199–225). OECD Publishing.
Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed-effects models using
lme4. Journal of Statistical Software, 67(1), 1–48. http://doi.org/10.18637/jss.v067.i01
Baumert, J., & Kunter, M. (2006). Stichwort: Professionelle Kompetenz von Lehrkräften.
[Keyword: Professional competence of teachers.]. Zeitschrift für Erziehungswissenschaft, 9(4),
469–520. https://doi.org/10.1007/s11618-006-0165-2
 INTERNATIONAL JOURNAL OF SCIENCE EDUCATION 19

Behrendt, H. (1991). Physikalische Schulversuche: Didaktische Theorie, methodische Praxis und die
Einstellung von Schülern zur Auswahl der Versuchsgeräte. [School physics experiments: didactic
theory, methodological practice, and student attitudes toward experimental equipment selec­
tion.] Kiel: PH Kiel.
Berg, C. A. R., Bergendahl, V. C. B., Lundberg, B., & Tibell, L. (2003). Benefiting from an open-
ended experiment? A comparison of attitudes to, and outcomes of, an expository versus
open-inquiry version of the same experiment. International Journal of Science Education,
25(3), 351–372. https://doi.org/10.1080/09500690210145738
Blanchard, M. R., Southerland, S. A., Osborne, J. W., Sampson, V. D., Annetta, L. A., & Granger, E.
M. (2010). Is inquiry possible in light of accountability? A quantitative comparison of the rela­
tive effectiveness of guided inquiry and verification laboratory instruction. Science Education,
94, 577–616. https://doi.org/10.1002/sce.20390
Bransford, J. D., Brown, A. L., & Cocking, R. R. (2000). How people learn: Brain, mind, experience,
and school. National Academy Press.
Caglak, S. (2017). Does hands-on science practices make an impact on achievement in science? A
meta-analysis. Journal of Education in Science, Environment and Health, 3(1), 69–87. https://doi.
org/10.21891/jeseh.275708
Cella, D., Gershon, R., Lai, J.-S., & Choi, S. (2007). The future of outcomes measurement: Item
banking, tailored short-forms, and computerized adaptive assessment. Quality of Life
Research, 16(Suppl 1), 133–141. https://doi.org/10.1007/s11136-007-9204-6
Cerini, B., Murray, I., & Reiss, M. (2003). Student review of the science curriculum. Major findings.
Planet Science/Institute of Education University of London/Science Museum.
Collins, A., Brown, J. S., & Newman, S. E. (1989). Cognitive apprenticeship: Teaching the craft of
Reading, writing, and mathematics. In L. B. Resnick (Ed.), Knowing, learning and instruction:
Essays in honor of robert glaser (pp. 453–494). Erlbaum.
Crawford, B. A. (2014). From inquiry to scientific practices in the science classroom. In N. G.
Lederman, & S. K. Abell (Eds.), Handbook of research on science education (Vol. II, pp. 515–
541). Routledge.
De Jong, D., Van Joolingen, T., & R, W. (1998). Scientific discovery learning with computer simu­
lations of conceptual domains. Review of Educational Research, 68(2), 179–201. https://doi.org/
10.3102/00346543068002179
Dekkers, P. J. J. M., & Thijs, G. D. (1998). Making productive use of students’ initial conceptions in
developing the concept of force. Science Education, 32(1), 31–51.
de Winter, J., & Airey, J. (2020). What makes a good physics teacher? Views from the English sta­
keholder community. Physics Education, 55(1), 15017. https://doi.org/10.1088/1361-6552/
ab5215
de Winter, J., & Millar, R. (2023). From broad principles to content-specific decisions: Pre-service
physics teachers’ views on the usefulness of practical work. International Journal of Science
Education, 45(13), 1097–1117. https://doi.org/10.1080/09500693.2023.2187673
Dillon, J. (2008). A review of the research on practical work in school science. King’s College.
Dobber, M., Zwart, R., Tanis, M., & van Oers, B. (2017). Literature review. The role of the teacher
in inquiry-based education. Educational Research Review, 22, 94–214. https://doi.org/10.1016/j.
edurev.2017.09.002
Furtak, E. M., & Penuel, W. R. (2019). Coming to terms: Addressing the persistence of “hands-on”
and other reform terminology in the era of science as practice. Science Education, 103(1), 167–
186. https://doi.org/10.1002/sce.21488
Furtak, E. M., Seidel, T., Iverson, H., & Briggs, D. C. (2012). Experimental and quasi-experimental
studies of inquiry-based science teaching. Review of Educational Research, 82(3), 300–329.
https://doi.org/10.3102/0034654312457206
Furtak, E. M., Shavelson, R. J., Shemwell, J. T., & Figueroa, M. (2012a). To teach or not to teach
through inquiry. In J. Shrager, & S. Carver (Eds.), The journey from child to scientist: Integrating
cognitive development and the education sciences (pp. 227–244). APA. https://doi.org/10.1037/
13617-011
20 J. WINKELMANN ET AL.

Gericke, N., Högström, P., & Wallin, J. (2022). A systematic review of research on laboratory work
in secondary school. Studies in Science Education, 59(2), 245–285. https://doi.org/10.1080/
03057267.2022.2090125
Grund, S., Robitzsch, A., & Luedtke, O. (2021). mitml: Tools for multiple imputation in multilevel
modeling (0.4.3).
Hacking, I. (1983). Representing and intervening: Introductory topics in the philosophy of natural
science. Harvard University Press.
Halawa, S., Hsu, Y.-S., Zhang, W.-X., Kuo, Y.-R., & Wu, J.-Y. (2020). Features and trends of teach­
ing strategies for scientific practices from a review of 2008-2017 articles. International Journal of
Science Education, 42(7), 1183–1206. https://doi.org/10.1080/09500693.2020.1752415
Hart, C., Mulhall, P., Berry, A., Loughran, J., & Gunstone, R. (2000). What is the purpose of this
experiment? Or can students learn something from doing experiments? Journal of Research in
Science Teaching, 37(7), 655–675. https://doi.org/10.1002/1098-2736(200009)37:7
Hattie, J. A. C. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to achieve­
ment. Routledge.
Hedges, L. V., & Hedberg, E. C. (2013). Intraclass correlations and covariate outcome correlations
for planning two-and three-level cluster-randomized experiments in education. Evaluation
Review, 37(6), 445–489. https://doi.org/10.1177/0193841X14529126
Heindl, M. (2019). Inquiry-based learning and the pre-requisite for its use in science at school: A
meta-analysis. Journal of Pedagogical Research, 3(2), 52–61. https://doi.org/10.33902/JPR.
2019254160
Hmelo-Silver, C. E., Duncan, R. G., & Chinn, C. A. (2007). Scaffolding and achievement in
problem-based and inquiry learning. Educational Psychologist, 42(2), 99–107. https://doi.org/
10.1080/00461520701263368
Hodson, D., & Bencze, L. (1998). Becoming critical about practical work: Changing views and
changing practice through action research. International Journal of Science Education, 20(6),
683–694. https://doi.org/10.1080/0950069980200606
Hofstein, A., & Lunetta, V. N. (1982). The role of the laboratory in science teaching: Neglected
aspects of research. Review of Educational Research, 52(2), 201–217. https://doi.org/10.3102/
00346543052002201
Hofstein, A., & Lunetta, V. N. (2003). The laboratory in science education: Foundations for the
twenty-first century. Science & Education, 88, 28–54.
Högström, P., Ottander, C., & Benckert, S. (2010). Lab work and learning in secondary school
chemistry. The importance of teacher and student interaction. Research in Science Education,
40, 505–523. https://doi.org/10.1007/s11165-009-9131-3
Hood Cattaneo, K. (2017). Telling active learning pedagogies apart. Journal of New Approaches in
Educational Research, 6(2), 144–152. https://doi.org/10.7821/naer.2017.7.237
Hopf, M. (2007). Problemorientierte Schülerexperimente. [Problem-oriented practical work]. In
H. Niedderer, H. Fischler, & E. Sumfleth (Eds.), Studien zum Physik- und Chemielernen.
[Studies of physics and chemistry education] (Vol. 68). Logos.
Hox, J. J. (2002). Multilevel analysis: Techniques and applications. Erlbaum.
Hox, J. J., Moerbeek, M., & Van de Schoot, R. (2017). Multilevel analysis: Techniques and appli­
cations. Routledge.
Jerrim, J., Oliver, M., & Sims, S. (2019). The relationship between inquiry-based teaching and stu­
dents’ achievement. New evidence from a longitudinal PISA study in England. Learning and
Instruction, 6, 35–44. https://doi.org/10.1016/j.learninstruc.2018.12.004
Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during instruction does
Not work: An analysis of the failure of constructivist, discovery, problem-based, experiential,
and inquiry-based teaching. Educational Psychologist, 41(2), 75–86. https://doi.org/10.1207/
s15326985ep4102_1
Klahr, D., Zimmerman, C., & Jirout, J. (2011). Educational interventions to advance children’s
scientific thinking. Science, 333, 971–975. http://doi.org/10.1126/science.1204528
 INTERNATIONAL JOURNAL OF SCIENCE EDUCATION 21

Kunter, M., Klusmann, U., Baumert, J., Richter, D., Voss, T., & Hachfeld, A. (2013). Professional
competence of teachers: Effects on instructional quality and student development. Journal of
Educational Psychology, 105(3), 805–820. https://doi.org/10.1037/a0032583
Lamprecht, J. (2011). Ausbildungswege und Komponenten professioneller Handlungskompetenz.
[Educational pathways and components of professional competence.]. In H. Niedderer, H.
Fischler, & E. Sumfleth (Eds.), Studien zum Physik- und Chemielernen [Studies of physics and
chemistry education] (Vol. 125). Logos.
Lazonder, A. W., & Harmsen, R. (2016). Meta-analysis of inquiry-based learning: Effects of gui­
dance. Review of Educational Research, 86(3), 681–718. https://doi.org/10.3102/
0034654315627366
Lederman, N. G., Antink, A., & Bartos, S. (2014). Nature of science, scientific inquiry, and socio-
scientific issues arising from genetics: A pathway to developing a scientifically literate citizenry.
Science & Education, 23(2), 285–302. https://doi.org/10.1007/s11191-012-9503-3
Li, K. H., Raghunathan, T. E., & Rubin, D. B. (1991). Large-sample significance levels from mul­
tiply imputed data using moment-based statistics and an F reference distribution. Journal of the
American Statistical Association, 86(416), 1065–1073. https://doi.org/10.1080/01621459.1991.
10475152
Lüdtke, O., Robitzsch, A., Trautwein, U., & Köller, O. (2007). Umgang mit fehlenden Werten in
der psychologischen Forschung. [Dealing with missing values in psychological research].
Psychologische Rundschau, 58(2), 103–117. https://doi.org/10.1026/0033-3042.58.2.103
Lunetta, V. N., Hofstein, A., & Clough, M. P. (2007). Teaching and learning in the school science
laboratory: An analysis of research, theory, and practice. In S. K. Abell, & N. G. Lederman
(Eds.), Handbook of research on science education (pp. 393–431). Lawrence Erlbaum.
Maričić, M., Cvjetićanin, S., & Anđić, B. (2019). Teacher-demonstration and student hands-on
experiments in teaching integrated sciences. Journal of Baltic Science Education, 18(5), 768–
779. https://doi.org/10.33225/jbse/19.18.768
Mayer, R. E. (2004). Should there be a three-strikes rule against pure discovery learning? The
American Psychologist, 59(1), 14–19. https://doi.org/10.1037/0003-066X.59.1.14
Millar, R. (2010). Practical work. In J. Osborne, & J. Dillon (Eds.), Good practice in science teaching:
What research has to say (pp. 108–134). Open University Press.
Newman, D. A. (2014). Missing data. Organizational Research Methods, 17(4), 372–411. https://
doi.org/10.1177/1094428114548590
NGSS, Next Generation Science Standards. (2013). Published on https://www.nextgenscience.org,
retrieved on 14.2.2022
NRC, National Research Council. (1996). National science education standards. National
Academies Press.
NRC, National Research Council. (2012). A framework for K-12 science education. Practices, cross­
cutting concepts, and core ideas. National Academies Press.
Oettinghaus, L., Lamprecht, J., & Korneck, F. (2012). Überzeugungen zum Unterrichtsfach und
zur Wissenschaft Physik - Ein Skalenvergleich. [Beliefs about the subject and science of
physics - A scale comparison.] PhyDid B. https://ojs.dpg-physik.de/index.php/phydid-b/
article/view/397/493
Osborne, J. (2014). Teaching scientific practices: Meeting the challenge of change. Journal of
Science Teacher Education, 25(2), 177–196. https://doi.org/10.1007/s10972-014-9384-1
Osborne, J. (2019). Not “hands on” but “minds on”: A response to Furtak and Penuel. Science
Education, 103(5), 1280–1283. https://doi.org/10.1002/sce.21543
Osborne, J., Simon, S., & Collins, S. (2003). Attitudes towards science: A review of the literature
and its implications. International Journal of Science Education, 25, 1049–1079. https://doi.
org/10.1080/0950069032000032199
Peugh, J. L., & Enders, C. K. (2004). Missing data in educational research: A review of reporting
practices and suggestions for improvement. Review of Educational Research, 74(4), 525–556.
https://doi.org/10.3102/00346543074004525
22 J. WINKELMANN ET AL.

Potvin, P., & Hasni, A. (2014). Interest, motivation and attitude towards science and technology at
K-12 levels: A systematic review of 12 years of educational research. Studies in Science
Education, 50(1), 85–129. https://doi.org/10.1080/03057267.2014.881626
Potvin, P., Hasni, A., & Sy, O. (2017). Using inquiry-based interventions to improve secondary
students’ interest in science and technology. European Journal of Science and Mathematics
Education, 5(3), 262–270. https://doi.org/10.30935/scimath/9510
R Core Team. (2021). R: A language and environment for statistical computing. (Version 4.1.1). R
Foundation for Statistical. http://www.R-project.org/
Rönnebeck, S., Bernholt, S., & Ropohl, M. (2016). Searching for a common ground: A literature
review of empirical research on scientific inquiry activities. Studies in Science Education,
52(2), 161–197. https://doi.org/10.1080/03057267.2016.1206351
Rumann, S. (2005). Kooperatives Experimentieren im Chemieunterricht: Entwicklung und
Evaluation einer Interventionsstudie zur Säure-Base-Thematik [Cooperative practical work in
school chemistry: Development and evaluation of an intervention study on the topic of acid
and base]. In H. Niedderer, H. Fischler, & E. Sumfleth (Eds.), Studien zum Physik- und
Chemielernen [Studies of physics and chemistry education] (Vol. 45). Logos.
Schmidt, H., Loyens, S. M. M., van Gog, T., & Paas, F. (2007). Problem-Based learning is compa­
tible with human cognitive architecture: Commentary on Kirschner, Sweller, and lark (2006).
Educational Psychologist, 42(2), 91–97. https://doi.org/10.1080/00461520701263350
Schreiber, N., Theyßen, H., & Schecker, H. (2009). Experimentelle Kompetenz messen?! [Testing
competencies of doing practical work?!]. Physik und Didaktik in Schule und Hochschule, 3(8),
92–101.
Schulz, A. (2011). Experimentierspezifische Qualitätsmerkmale im Chemieunterricht: Eine
Videostudie [Major quality characteristics for practical work in school chemistry: A video
study]. In H. Niedderer, H. Fischler, & E. Sumfleth (Eds.), Studien zum Physik- und
Chemielernen [studies of physics and chemistry education] (Vol. 113). Logos.
Schwichow, M., Zimmerman, C., Croker, S., & Härtig, H. (2016). What students learn from hands-
on activities. Journal of Research in Science Teaching, 1–23.
Seidel, T., Schwindt, K., Rimmele, R., & Prenzel, M. (2008). Konstruktivistische Überzeugungen
von Lehrpersonen: Was bedeuten sie für den Unterricht? [teachers’ constructivist beliefs:
What do they mean for teaching?] ZfE. Sonderheft, 9, 259–276.
Seidel, T., & Shavelson, R. J. (2007). Teaching effectiveness research in the past decade: The role of
theory and research design in disentangling meta-analysis results. Review of Educational
Research, 77(4), 454–499. https://doi.org/10.3102/0034654307310317
Shana, Z., & Abulibdeh, E. S. (2020). Science practical work and its impact on students’ science
achievement. Journal of Technology and Science Education, 10(2), 199–215. https://doi.org/10.
3926/jotse.888
Shulman, L. S., & Tamir, P. (1973). Research on teaching in the natural sciences. In R. M. W.
Travers (Ed.), Second handbook of research on teaching (pp. 1098–1148). Rand-McNally.
Singer, S. R., Hilton, M. L., & Schweingruber, H. A. (2006). America’s lab report. Investigations in
high school science. Washington, DC: The National Academies. https://doi.org/10.17226/11311
Skott, J. (2015). The promises, problems, and prospects of research on teachers’ beliefs. In H. Fives
(Ed.), International handbook of research on teachers’ beliefs (S. 13–29). Routledge.
Staub, F. C., & Stern, E. (2002). The nature of teachers’ pedagogical content beliefs matters for stu­
dents’ achievement gains: Quasi-experimental evidence from elementary mathematics. Journal
of Educational Psychology, 94(2), 344–355. https://doi.org/10.1037/0022-0663.94.2.344
Stipek, D. J., Givvin, K. B., Salmon, J. M., & MacGyvers, V. L. (2001). Teachers’ beliefs and prac­
tices related to mathematics instruction. Teaching and Teacher Education, 17(2), 213–226.
https://doi.org/10.1016/S0742-051X(00)00052-4
Tesch, M. (2005). Das Experiment im Physikunterricht: Didaktische Konzepte und Ergebnisse
einer Videostudie [practical work in school physics: Didactically concepts and results of a
video study]. In H. Niedderer, H. Fischler, & E. Sumfleth (Eds.), Studien zum Physik- und
Chemielernen [studies of physics and chemistry education] (Vol. 42). Logos.
 INTERNATIONAL JOURNAL OF SCIENCE EDUCATION 23

Trumper, R. (2003). The physics laboratory. A historical overview and future perspectives. Science
and Education, 12, 645–670. https://doi.org/10.1023/A:1025692409001
van Buuren, S. (2018). Flexible imputation of missing data, second edition. Chapman and Hall/
CRC. https://doi.org/10.1201/9780429492259
Velasco, J. B., Knedeisen, A., Xue, D., Vickrey, T. L., Abebe, M., & Stains, M. (2016).
Characterizing instructional practices in the laboratory: The laboratory observation protocol
for undergraduate STEM. Journal of Chemical Education, 93(7), 1191–1203. https://doi.org/
10.1021/acs.jchemed.6b00062
von Davier, M., & Sinharay, S. (2014). Analytics in international large-scale assessments: Item
response theory and population models. In L. Rutkowski, M. von Davier, & D. Rutkowski
(Eds.), Handbook of international large-scale assessment. Background, technical issues, and
methods of data analysis (pp. 155–174). Routledge.
Voss, T., Kleickmann, T., Kunter, M., & Hachfeld, A. (2013). Mathematics teachers’ beliefs. In M.
Kunter, J. Baumert, W. Blum, U. Klusmann, S. Krauss, & M. Neubrand (Eds.), Cognitive acti­
vation in the mathematics classroom and professional competence of teachers. Results from the
COACTIV project (pp. 249–271). Springer.
Wallace, C. S. (2014). Overview of the role of teacher beliefs in science education. In R. Evans, J.
Luft, C. Czerniak, & C. Pea (Eds.), The role of science teachers’ beliefs in international classrooms
(pp. 17–31). Sense Publishers.
Weber, J., Wenzel, F., Winkelmann, J., Ullrich, M., Horz, H., & Erb, R. (2018). Subject knowledge
in geometrical optics: Testing and improving student’s knowledge. In O. E. Finlayson, E.
McLoughlin, S. Erduran, & P. Childs (Eds.), Electronic proceedings of the ESERA 2017 confer­
ence. Research, practice and collaboration in science education, part 3 (co-ed. Sabine Fechner
& Andrée Tiberghien) (pp. 448–455). Dublin City University.
Welzel, M., Haller, K., Bandiera, M., Hammelev, D., Koumaras, P., Niedderer, H., & von
Aufschnaiter, S. (1998). Ziele, die Lehrende mit dem Experimentieren in der naturwissenschaf­
tlichen Ausbildung verbinden. [Goals Teachers Associate with Experimenting in Science
Education]. Zeitschrift für Didaktik der Naturwissenschaften, 4(1), 29–44.
Winkelmann, J. (2015). Auswirkungen auf den Fachwissenszuwachs und auf affektive
Schülermerkmale durch Schüler- und Demonstrationsexperimente im Physikunterricht.
[Effects on subject knowledge growth and affective student characteristics of student and dem­
onstration experiments in physics classrooms.]. In H. Niedderer, H. Fischler, & E. Sumfleth
(Eds.), Studien zum Physik- und Chemielernen. [studies of physics and chemistry education]
(Vol. 179). Logos Verlag.
Winkelmann, J., & Erb, R. (2011). Auswirkungen auf den Fachwissenszuwachs und auf affektive
Schülermerkmale durch Schüler- und Demonstrationsexperimente im Physikunterricht. In H.
Grötzebauch, & V. Nordmeier (Eds.), Phydid B – Didaktik der Physik: Beiträge zur DPG-
Frühjahrstagung des Fachverbands Didaktik der Physik in Münster 2011, Paper no. 15.02. DPG.
Winkelmann, J., & Erb, R. (2018). Der Einfluss von Schüler- und Demonstrationsexperimenten
auf den Lernzuwachs in Physik. [The impact of student and demonstration experiments on
learning gains in physics.] PhyDid A. Physik und Didaktik in Schule und Hochschule, 17(1),
21–33.
Yang, X., Kaiser, G., Konig, J., & Blomeke, S. (2020). Relationship between pre-service mathemat­
ics teachers’ knowledge, beliefs and instructional practices in China. ZDM Mathematics
Education, 52(2), 281–294. https://doi.org/10.1007/s11858-020-01145-x