Download the Full Version of textbook for Fast Typing at textbookfull.

com

Psychophysics, Second Edition: A Practical

Introduction Kingdom

https://textbookfull.com/product/psychophysics-second-
edition-a-practical-introduction-kingdom/

OR CLICK BUTTON

DOWNLOAD NOW

Download More textbook Instantly Today - Get Yours Now at textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Biota Grow 2C gather 2C cook Loucas

https://textbookfull.com/product/biota-grow-2c-gather-2c-cook-loucas/

textboxfull.com

ENT: An Introduction and Practical Guide, Second Edition

James Tysome

https://textbookfull.com/product/ent-an-introduction-and-practical-
guide-second-edition-james-tysome/

textboxfull.com

Modern statistics for the social and behavioral sciences :

a practical introduction Second Edition. Edition Wilcox

https://textbookfull.com/product/modern-statistics-for-the-social-and-
behavioral-sciences-a-practical-introduction-second-edition-edition-
wilcox/
textboxfull.com

Kingdom Ethics Following Jesus in Contemporary Context

Second Edition Gushee

https://textbookfull.com/product/kingdom-ethics-following-jesus-in-
contemporary-context-second-edition-gushee/

textboxfull.com
Computational seismology : a practical introduction 1st
Edition Igel

https://textbookfull.com/product/computational-seismology-a-practical-
introduction-1st-edition-igel/

textboxfull.com

Management: A Practical Introduction 7th Edition Angelo

Kinicki

https://textbookfull.com/product/management-a-practical-
introduction-7th-edition-angelo-kinicki/

textboxfull.com

Literary theory a practical introduction Third Edition

Rivkin

https://textbookfull.com/product/literary-theory-a-practical-
introduction-third-edition-rivkin/

textboxfull.com

Discourse Analysis: A Practical Introduction 1st Edition

Canning

https://textbookfull.com/product/discourse-analysis-a-practical-
introduction-1st-edition-canning/

textboxfull.com

Discourse Analysis: A Practical Introduction 1st Edition

Canning

https://textbookfull.com/product/discourse-analysis-a-practical-
introduction-1st-edition-canning-2/

textboxfull.com
PSYCHOPHYSICS
A PRACTICAL INTRODUCTION

SECOND EDITION
FREDERICK A.A. KINGDOM
McGill University, Montreal, Quebec, Canada

NICOLAAS PRINS
University of Mississippi, Oxford, MS, USA

AMSTERDAM • BOSTON • HEIDELBERG • LONDON

NEW YORK • OXFORD • PARIS • SAN DIEGO
SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO
Academic Press is an imprint of Elsevier
Academic Press is an imprint of Elsevier
125 London Wall, London EC2Y 5AS, UK
525 B Street, Suite 1800, San Diego, CA 92101-4495, USA
225 Wyman Street, Waltham, MA 02451, USA
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK

Copyright © 2016, 2010 Elsevier Ltd. All rights reserved.

Cover image: This item is reproduced by permission of The Huntington Library, San Marino, California.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical,
Including photocopying, recording, or any information storage and retrieval system, without permission in writing from
the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our
arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be
found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than
as may be noted herein).

Notices
Knowledge and best practice in this field are constantly changing. As new research and experience broaden our
understanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any
information, methods, compounds, or experiments described herein. In using such information or methods they should
be mindful of their own safety and the safety of others, including parties for whom they have a professional
responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for
any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any
use or operation of any methods, products, instructions, or ideas contained in the material herein.

ISBN: 978-0-12-407156-8

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

For information on all Academic Press publications

visit our website at http://store.elsevier.com/

Publisher: Mica Haley

Acquisition Editor: Melanie Tucker
Editorial Project Manager: Kristi Anderson
Production Project Manager: Caroline Johnson
Designer: Matt Limbert

Typeset by TNQ Books and Journals

www.tnq.co.in

Printed and bound in the United States of America

 Dedication

FK would like to dedicate this book to his late parents Tony and Joan, and present
family Beverley and Leina. NP would like to dedicate this book to his mother Nel and late
father Arie.
 About the Authors

Frederick A.A. Kingdom is a Professor at McGill University conducting research into

various aspects of visual perception, including color vision, brightness perception, stereopsis,
texture perception, contour-shape coding, the perception of transparency, and visual illu-
sions. He also has an interest in models of summation for the detection of multiple stimuli.
Nicolaas Prins is an Associate Professor at the University of Mississippi specializing in
visual texture perception, motion perception, contour-shape coding, and the use of statistical
methods in the collection and analysis of psychophysical data.

ix
 Preface to the Second Edition

The impetus for this book was a recurring comparisons. We have also provided an
question: “Is there a book that explains how updated quick reference guide to the terms,
to do psychophysics?” Evidently, a book was concepts, and many of the equations
needed that not only explained the theory described in the book.
behind psychophysical procedures but also In writing the second edition we have
provided the practical tools necessary for endeavored to improve each chapter and
their implementation. What seemed to be have extended all the technical chapters to
missing was a detailed and accessible expo- include new procedures and analyses.
sition of how raw psychophysical responses Chapter 7 is the book’s one new chapter. It
are turned into meaningful measurements of deals with an old but vexing question of
sensory function; in other words, a book that how multiple stimuli combine to reach
dealt with the nuts and bolts of psycho- threshold. The chapter attempts to derive
physics data analysis. from first principles and make accessible to
The need for a practical book on psycho- the reader the mathematical basis of the
physics inevitably led to a second need: a myriads of summation models, scenarios,
comprehensive package of software for and metrics that are scattered throughout
analyzing psychophysical data. The result the literature.
was Palamedes. Initially developed in Writing both editions of this book has
conjunction with the first edition of the book, been a considerable challenge for its authors.
Palamedes has since taken on a life of its Much effort has been expended in trying to
own, and one purpose of the second edition make accessible the theory behind different
is to catch up with its latest developments! types of psychophysical data analysis. For
Palamedes will of course continue to be those psychophysical terms that to us did
developed so readers are encouraged to keep not appear to have a clear definition we have
an eye on the regular updates. improvised our own (e.g., the definition of
The first few chapters of the book are “appearance” given in Chapter 2), and for
intended to introduce the basic concepts and other terms where we felt there was a lack of
terminology of psychophysics as well as clarity we have challenged existing conven-
familiarize readers with a range of psycho- tion (e.g., by referring to a class of forced-
physical procedures. The remaining chapters choice tasks as 1AFC). Where we have
focus on specialist topics: psychometric challenged convention we have explained
functions, adaptive procedures, signal our reasoning and hope that even if readers
detection theory, summation measures, do not agree with us, they will still find our
scaling methods, and statistical model ideas on the matter thought-provoking.

xi
 Acknowledgments

We are indebted to the following persons for kindly reviewing and providing insightful
comments on individual chapters: Neil Macmillan and Douglas Creelman for helping one of
the authors (FK) get to grips with the calculation of d0 for same-different tasks (Chapter 6);
Mark Georgeson for providing the derivation of the equation for the criterion measure lnb for
a 2AFC task (Chapter 6); Alex Baldwin for the idea of incorporating a stimulus scaling factor
g for converting stimulus intensity to d0 when modeling psychometric functions within a
Signal Detection Theory framework (Chapters 6 and 7); Mark McCourt for providing the
figures illustrating grating-induction (Chapter 3); Laurence Maloney for permission to
develop and describe the routines for Maximum Likelihood Difference Scaling (Chapter 8);
Stanley Klein for encouraging us to include a section on the Chi-squared test (Chapter 9); and
Ben Jennings for carefully checking the equations in the summation chapter (Chapter 7).
Thanks also to the many personsdtoo many to mention individuallydwho have over the
years expressed their appreciation for the book as well as the Palamedes toolbox and
provided useful suggestions for improvements to both.

xiii
 C H A P T E R

1
Introduction and Aims
Frederick A.A. Kingdom1, Nicolaas Prins2
1
McGill University, Montreal, Quebec, Canada; 2University of Mississippi, Oxford, MS, USA

O U T L I N E

1.1 What is Psychophysics? 1 1.4 What’s New in the Second

Edition? 5
1.2 Aims of the Book 1
References 9
1.3 Organization of the Book 2

1.1 WHAT IS PSYCHOPHYSICS?

According to the online encyclopedia Wikipedia, psychophysics “. quantitatively investi-

gates the relationship between physical stimuli and the sensations and perceptions they
affect.” The term was first coined by Gustav Theodor Fechner, who in his Elements of Psycho-
physics (1860/1966) set out the principles of psychophysical measurement, describing the
various procedures used by experimentalists to map out the relationship between matter
and mind. Although psychophysics refers to a methodology, it is also a research area in its
own right, and much effort continues to be devoted to developing new psychophysical tech-
niques and new methods for analyzing psychophysical data.
Psychophysics can be applied to any sensory system, whether vision, hearing, touch, taste,
or smell. This book primarily draws on the visual system to illustrate the principles of
psychophysics, but the principles are applicable to all sensory domains.

1.2 AIMS OF THE BOOK

Broadly speaking, the book has three aims. The first is to provide newcomers to psycho-
physics with an overview of different psychophysical procedures in order to help them

Psychophysics
http://dx.doi.org/10.1016/B978-0-12-407156-8.00001-3 1 Copyright © 2016 Elsevier Ltd. All rights reserved.
2 1. INTRODUCTION AND AIMS

select the appropriate designs and analyses for their experiments. The second aim is to
direct readers to the software tools, in the form of Palamedes, for analyzing psychophysical
data. This is intended for both newcomers and experienced researchers alike. The third aim
is to explain the theory behind the analyses. Again both newcomers and experienced re-
searchers should benefit from the detailed expositions of the bulk of the underlying theory.
To this end we have made every effort to make accessible the theory behind a wide range of
psychophysical procedures, analytical principles, and mathematical computations, such as
Bayesian curve fitting; the calculation of d-primes (dʹ); summation theory; maximum likeli-
hood difference scaling; goodness-of-fit measurement; bootstrap analysis; and likelihood-
ratio testing, to name but a few. In short, the book is intended to be both practical and
pedagogical.
The inclusion of the description of the Palamedes tools, placed in this edition in
separate boxes alongside the main text, will hopefully offer the reader something more
than is provided by traditional textbooks, such as the excellent Psychophysics: The Funda-
mentals by Gescheider (1997). If there is a downside, however, it is that we do not always
delve as deeply into the relationship between psychophysical measurement and sensory
function as The Fundamentals does, except when necessary to explain a particular psycho-
physical procedure or set of procedures. In this regard A Practical Introduction is not
intended as a replacement for other textbooks on psychophysics but as a complement to
them, and readers are encouraged to read other relevant texts alongside our own. Two
noteworthy recent additions to the literature on psychophysics are Knoblauch and
Maloney’s (2012) Modeling Psychophysical Data in R and Lu and Dosher’s (2013) Visual
Psychophysics.
Our approach of combining the practical and the pedagogical into a single book may not
be to everyone’s taste. Doubtless some would prefer to have the description of the software
routines put elsewhere. However, we believe that by describing the software alongside the
theory, newcomers will be able to get a quick handle on the nuts and bolts of
psychophysics methods, the better to then delve into the underlying theory if and when
they choose.

1.3 ORGANIZATION OF THE BOOK

The book can be roughly divided into two parts. Chapters 2 and 3 provide an overall
framework and detailed breakdown of the variety of psychophysical procedures available
to the researcher. Chapters 4e9 are the technical chapters. They describe the theory and
implementation for six specialist topics: psychometric functions; adaptive methods;
signal detection measures; summation measures; scaling methods; and model comparisons
(Box 1.1).
In Chapter 2 we provide an overview of some of the major varieties of psychophysical
procedures and offer a framework for classifying psychophysics experiments. The approach
taken here is an unusual one. Psychophysical procedures are discussed in the context of a crit-
ical examination of the various dichotomies commonly used to differentiate psychophysics
experiments: Class A versus Class B; Type 1 versus Type 2; performance versus appearance;
forced-choice versus nonforced-choice; criterion-dependent versus criterion-free; objective
 1.3 ORGANIZATION OF THE BOOK 3

BOX 1.1

PALAMEDES
According to Wikipedia, the Greek mythological figure Palamedes (“pal-uh-MEE-deez”) is
said to have invented “. counting, currency, weights and measures, jokes, dice and a fore-
runner of chess called pessoi, as well as military ranks.” The story goes that Palamedes also
uncovered a ruse by Odysseus. Odysseus had promised Agamemnon that he would defend
the marriage of Helen and Menelaus but pretended to be insane to avoid having to honor his
commitment. Unfortunately, Palamedes’s unmasking of Odysseus led to a gruesome end; he
was stoned to death for being a traitor after Odysseus forged false evidence against him.
Palamedes was chosen as the name for the toolbox because of the legendary figure’s (pre-
sumed) contributions to the art of measurement, interest in stochastic processes (he did invent
dice!), numerical skills, humor, and wisdom. The Palamedes Swallowtail butterfly (Papilio
palamedes) on the front cover also provides the toolbox with an attractive icon.
Palamedes is a set of routines and demonstration programs written in MATLABÒ for
analyzing psychophysical data (Prins and Kingdom, 2009). The routines can be downloaded
from www.palamedestoolbox.org. We recommend that you check the website periodically,
because new and improved versions of the toolbox will be posted there for download.
Chapters 4e9 explain how to use the routines and describe the theory behind them. The
descriptions of Palamedes do not assume any knowledge of MATLAB, although a basic
knowledge will certainly help. Moreover, Palamedes requires only basic MATLAB; the
specialist toolboxes such as the Statistics toolbox are not required. We have also tried to make
the routines compatible with earlier versions of MATLAB, where necessary including alter-
native functions that are called when later versions are undetected. Palamedes is also
compatible with the free software package GNU Octave (http://www.octave.org).
It is important to bear in mind what Palamedes is not. It is not a package for generating
stimuli or for running experiments. In other words it is not a package for dealing with the
“front-end” of a psychophysics experiment. The two exceptions to this rule are the Palamedes
routines for adaptive methods, which are designed to be incorporated into an actual experi-
mental program, and the routines for generating stimulus lists for use in scaling experiments.
But by and large, Palamedes is a different category of toolbox from the stimulus-generating
toolboxes such as VideoToolbox (http://vision.nyu.edu/VideoToolbox/), PsychToolbox
(http://psychtoolbox.org), PsychoPy (http://www.psychopy.org; see also Peirce, 2007, 2009),
and Psykinematix (http://psykinematix.kybervision.net/) (for a comprehensive list of such
toolboxes see http://visionscience.com/documents/strasburger/strasburger.html). Although
some of these toolboxes contain routines that perform similar functions to some of the routines
in Palamedes, for example fitting psychometric functions (PFs), they are in general comple-
mentary to, rather than in competition with, Palamedes.
A few software packages deal primarily with the analysis of psychophysical data. Most of
these are aimed at fitting and analyzing psychometric functions. psignifit (http://psignifit.
sourceforge.net/; see also Fründ et al., 2011) is perhaps the best known of these. Another
option is quickpsy, written for R by Daniel Linares and Joan López-Moliner (http://dlinares.
org/quickpsy.html; see also Linares & López-Moliner, in preparation). Each of the packages

Continued
4 1. INTRODUCTION AND AIMS

BOX 1.1 (cont'd)

will have their own strengths and weaknesses and readers are encouraged to find the software
that best fits their needs. A major advantage of Palamedes is that it can fit PFs to multiple
conditions simultaneously, while providing the user considerable flexibility in defining a
model to fit. Just to give one simple example, one might assume that the lapse rate and slope of
the PF are equal between several conditions but that thresholds are not. Palamedes allows one
to specify and implement such assumptions and fit the conditions accordingly. Users can also
provide their own custom-defined relationships among the parameters from different condi-
tions. For example, users can specify a model in which threshold estimates in different
conditions adhere to an exponential decay function (or any other user-specified parametric
curve). Palamedes can also determine standard errors for the parameters estimated in such
multiple condition fits and perform goodness-of-fit tests for such fits.
The flexibility in model specification provided by Palamedes can also be used to perform
statistical model comparisons that target very specific research questions that a researcher
might have. Examples are to test whether thresholds differ significantly between two or more
conditions, to test whether it is reasonable to assume that slopes are equal between the con-
ditions, to test whether the lapse rate differs significantly from zero (or any other specific value),
to test whether the exponential decay function describes the pattern of thresholds well, etc.
Palamedes also does much more than fit PFs; it has routines for calculating signal detection
measures and summation measures, implementing adaptive procedures, and analyzing scaling
data.

versus subjective; detection versus discrimination; and threshold versus suprathreshold. We

consider whether any of these dichotomies could usefully form the basis of a fully-fledged
classification scheme for psychophysics experiments and conclude that one, the performance
versus appearance distinction, is the best candidate.
Chapter 3 takes as its starting point the classification scheme outlined in Chapter 2 and
expands on it by incorporating a further level of categorization based on the number of stim-
uli presented per trial. The expanded scheme serves as the framework for detailing a much
wider range of psychophysical procedures than described in Chapter 2.
Four of the technical chapters, Chapters 4, 6, 8, and 9, are divided into two sections. In
these chapters Section A introduces basic concepts and takes the reader through the Pala-
medes routines that perform the relevant data analyses. Section B provides more detail as
well as the theory behind the analyses. The idea behind the Section A versus Section B distinc-
tion is that readers can learn about the basic concepts and their implementation without
necessarily having to grasp the underlying theory, yet have the theory available to delve
into if they want. For example, Section A of Chapter 4 describes how to fit psychometric func-
tions and derive estimates of their critical parameters such as threshold and slope, while
Section B describes the theory behind the various fitting procedures. Similarly, Section A
 1.4 WHAT’S NEW IN THE SECOND EDITION? 5
in Chapter 6 outlines why dʹ measures are useful in psychophysics and how they can be
calculated using Palamedes, while Section B describes the theory behind the calculations.
Here and there, we present specific topics in some detail in separate boxes. The idea behind
this is that the reader can easily skip these boxes without loss of continuity, while readers spe-
cifically interested in the topics discussed will be able to find detailed information there. Just
to give one example, Box 4.6 in Chapter 4 explains in much detail the procedure that is used
to fit a psychometric function to some data, gives information as to how some fits might fail,
and provides tips on how to avoid failed fits.

1.4 WHAT’S NEW IN THE SECOND EDITION?

A major change from the first edition is the addition of the chapter on summation mea-
sures (Chapter 7). This chapter provides a detailed exposition of the theory and practice
behind experiments that measure detection thresholds for multiple stimuli. Besides the
new chapter, all the other chapters have been rewritten to a greater or lesser degree, mainly
to include new procedures and additional examples.
Another important change from the first edition is that the description of the Palamedes
routines has been put into boxes placed alongside the relevant text. This gives readers greater
flexibility in terms of whether, when, and where they choose to learn about Palamedes. The
boxes in this chapter (Box 1 through Box 3) are designed to introduce the reader to Palamedes
and its implementation in MATLAB.

BOX 1.2

ORGANIZATION OF PALAMEDES
All the Palamedes routines are prefixed by an identifier PAL, to avoid confusion with the
routines used by MATLAB. After PAL, many routine names contain an acronym for the class of
procedure they implement. Box 1.3 lists the acronyms currently in the toolbox, what they stand
for, and the book chapter where they are described. In addition to the routines with specialist
acronyms, there are a number of general-purpose routines.

Functions
In MATLAB there is a distinction between a function and a script. A function accepts one or
more input arguments, performs a set of operations, and returns one or more output argu-
ments. Typically, Palamedes functions are called as follows:
>>[x y z] ¼ PAL_FunctionName(a,b,c);

where a, b, and c are the input arguments, and x, y, and z the output arguments. In general,
the input arguments are “arrays.” Arrays are simply listings of numbers. A scalar is a single
number, e.g., 10, 1.5, 1.0ee15. A vector is a one-dimensional array of numbers. A matrix is a
two-dimensional array of numbers. It will help you to think of all as being arrays. As a matter
of fact, MATLAB represents all as two-dimensional arrays. That is, a scalar is represented as a

Continued
6 1. INTRODUCTION AND AIMS

BOX 1.2 (cont'd)

1  1 (1 row  1 column) array, vectors either as an m  1 array or a 1  n array, and a matrix

as an m  n array. Arrays can also have more than two dimensions.
In order to demonstrate the general usage of functions in MATLAB, Palamedes includes a
function named PAL_ExampleFunction, which takes two arrays of any dimensionality as
input arguments and returns the sum, the difference, the product, and the ratio of the numbers
in the arrays corresponding to the input arguments. For any function in Palamedes you can get
some information as to its usage by typing help followed by the name of the function:
>>help PAL_ExampleFunction

MATLAB returns
PAL_ExampleFunction calculates the sum, difference, product, and
ratio of two scalars, vectors or matrices.

syntax: [sum difference product ratio] ¼ ...

PAL_ExampleFunction(array1, array2)

This function serves no purpose other than to demonstrate the

general usage of Matlab functions.

For example, if we type and execute

[sum difference product ratio] ¼ PAL_ExampleFunction(10, 5);

MATLAB will assign the arithmetic sum of the input arguments to a variable labeled sum,
the difference to difference, etc. In case the variable sum did not previously exist, it will have
been created when the function was called. In case it did exist, its previous value will be
overwritten (and thus lost). We can inquire about the value of a variable by typing its name,
followed by <return>:
>>sum

MATLAB returns
sum ¼ 15

We can use any name for the returned arguments. For example, typing

>>[s d p r] ¼ PAL_ExampleFunction(10,5)

creates a variable s to store the sum, etc.

Instead of passing values directly to the function, we can assign the values to variables and
pass the name of the variables instead. For example the series of commands
>>a ¼ 10;
>>b ¼ 5;
>>[sum difference product ratio] ¼ PAL_ExampleFunction(a, b);
 1.4 WHAT’S NEW IN THE SECOND EDITION? 7

BOX 1.2 (cont'd)

generates the same result as before. You can also assign a single alphanumeric name to
vectors and matrices. For example, to create a vector called vect1 with values 1, 2, 4, and 105
one can simply type and follow with a <return>:
>> vect1 ¼ [1 2 4 105]

Note the square, not round brackets. vect1 can then be entered as an argument to a routine,
provided the routine is set up to accept a 1  4 vector. To create a matrix called matrix1
containing two columns and three rows of numbers, type and follow with a <return>, for
example
>> matrix1 ¼ [0.01 0.02; 0.04 0.05; 0.06 0.09]

where the semicolon separates the values for different rows. Again, matrix1 can now be
entered as an argument, provided the routine accepts a 3  2 (rows by columns) matrix.
Whenever a function returns more than one argument, we do not need to assign them all to
a variable. Let’s say we are interested in the sum and the difference of two matrices only. We
can type:
>>[sum difference] ¼ PAL_ExampleFunction([1 2; 3 4], [5 6; ...
7 8]);

Demonstration Programs
A separate set of Palamedes routines are suffixed by _Demo. These are located in the folder
PalamedesDemos separate from the other Palamedes routines. The files in the PalamedesDemos
folder are demonstration scripts that in general combine a number of Palamedes function
routines into a sequence to demonstrate some aspect of their combined operation. They pro-
duce a variety of types of output to the screen, such as numbers with headings, graphs, etc.
While these programs do not take arguments when they are called, the user might be
prompted to enter something when the program is run, e.g.,
>>PAL_Example_Demo
Enter a vector of stimulus levels

Then the user might enter something like [.1 .2 .3]. After pressing return there will be
some form of output, for example data with headings, a graph, or both.
Error Messages
The Palamedes toolbox is not particularly resistant to user error. Incorrect usage will more
often result in a termination of execution accompanied by an abstract error message than it will
in a gentle warning or a suggestion for proper usage. As an example, let us pass some

Continued
8 1. INTRODUCTION AND AIMS

BOX 1.2 (cont'd)

inappropriate arguments to our example function and see what happens. We will pass two
arrays to it of unequal size:
>>a ¼ [1 2 3];
>>b ¼ [4 5];
>>sum ¼ PAL_ExampleFunction(a, b);

MATLAB returns
??? Error using ¼¼> unknown
Matrix dimensions must agree.
Error in ¼¼> PAL_ExampleFunction at 15
sum ¼ array1 þ array2;

This is actually an error message generated by a resident MATLAB function, not a Pala-
medes function. Palamedes routines rely on many resident MATLAB functions and operators
(such as “þ”), and error messages you see will typically be generated by these resident
MATLAB routines. In this case, the problem arose when PAL_ExampleFunction attempted to
use the “þ”operator of MATLAB to add two arrays that are not of equal size.

BOX 1.3

ACRONYMS USED IN PALAMEDES

Acronyms used in names for Palamedes routines, their meaning, and the chapters in which they are
described

Acronym Meaning Chapter

AMPM Adaptive methods: psi method 5
AMRF Adaptive methods: running fit 5
AMUD Adaptive methods: up/down 5
MLDS Maximum likelihood difference scaling 7
PF Psychometric function 4
PFBA Psychometric function: Bayesian 4
PFLR Psychometric function: likelihood ratio 8
PFML Psychometric function: maximum likelihood 4, 8
SDT Signal detection theory 6
 REFERENCES 9

References
Fechner, G., 1860/1966. Elements of Psychophysics. Hilt, Rinehart & Winston, Inc.
Fründ, I., Haenel, N.V., Wichmann, F.A., 2011. Inference for psychometric functions in the presence of nonstationary
behavior. J. Vis. 11 (6), 16.
Gescheider, G.A., 1997. Psychophysics: The Fundamentals. Lawrence Erlbaum Associates, Mahwah, New Jersey.
Knoblauch, K., Maloney, L.T., 2012. Modeling Psychophysical Data in R. Springer.
Linares, D., López-Moliner, J., in preparation. Quickpsy: An R Package to Analyse Psychophysical Data.
Lu, Z.-L., Dosher, B., 2013. Visual Psychophysics. MIT Press, Cambridge, MA.
Peirce, J.W., 2007. PsychoPy e psychophysics software in Python. J. Neurosci. Methods 162 (1e2), 8e13.
Peirce, J.W., 2009. Generating stimuli for neuroscience using PsychoPy. Front. Neuroinform. 2, 10. http://dx.doi.org/
10.3389/neuro.11.010.2008.
Prins, N., Kingdom, F.A.A., 2009. Palamedes: MATLAB Routines for Analyzing Psychophysical Data. http://www.
palamedestoolbox.org.
 C H A P T E R

2
Classifying Psychophysical
Experiments*
Frederick A.A. Kingdom1, Nicolaas Prins2
1
McGill University, Montreal, Quebec, Canada; 2University of Mississippi, Oxford, MS, USA

O U T L I N E

2.1 Introduction 11 2.3.6 “Objective” versus “Subjective” 28

2.3.7 “Detection” versus
2.2 Tasks, Methods, and Measures 12
“Discrimination” 29
2.3 Dichotomies 14 2.3.8 “Threshold” versus
2.3.1 “Class A” versus “Class B” “Suprathreshold” 31
Observations 14
2.4 Classification Scheme 32
2.3.2 “Type 1” versus “Type 2” 19
2.3.3 “Performance” versus Further Reading 33
“Appearance” 20
Exercises 33
2.3.4 “Forced-Choice” versus
“Nonforced-Choice” 24 References 34
2.3.5 “Criterion-Free” versus
“Criterion-Dependent” 27

2.1 INTRODUCTION

This chapter describes various classes of psychophysical procedure and proposes a scheme
for classifying them. The aim is not so much to judge the pros and cons of different
proceduresdthis will be dealt with in the next chapterdbut to examine how they differ
and how they interrelate. The proposed classification scheme is arrived at through a critical

*
This chapter was primarily written by Frederick Kingdom.

Psychophysics
http://dx.doi.org/10.1016/B978-0-12-407156-8.00002-5 11 Copyright © 2016 Elsevier Ltd. All rights reserved.
12 2. CLASSIFYING PSYCHOPHYSICAL EXPERIMENTS

examination of the familiar “dichotomies” that make up the vernacular of psychophysics,

e.g., “Class A” versus “Class B” observations, “Type 1” versus “Type 2” tasks, “forced-
choice” versus “nonforced-choice” tasks, etc. These dichotomies do not always mean the
same thing to all people, so one of the aims of the chapter is to clarify what each dichotomy
means and consider how useful each might be as a category in a classification scheme.
Why a classification scheme? After all, the seasoned practitioner designs his or her psycho-
physics experiment based on knowledge accumulated over years of research experience,
including knowledge as to what is available, what is appropriate, and what is valid given
the question about visual function being asked. And that is how it should be. However, a
framework that captures both the critical differences as well as intimate relationships be-
tween different psychophysical procedures could be useful to newcomers in the field, helping
them to select the appropriate experimental design from what might seem a bewildering
array of possibilities. Thinking about a classification scheme is also a useful intellectual exer-
cise, not only for those of us who like to categorize things, put them into boxes, and attach
labels to them, but for anyone interested in gaining a deeper understanding of psychophysics.
But before discussing the dichotomies, consider the components that make up a psychophys-
ics experiment.

2.2 TASKS, METHODS, AND MEASURES

Although the outcome of a psychophysics experimentdtypically a set of measurementsd

reflects more than anything else the particular question about sensory function being
asked, other components of the experiment, in particular the stimulus and the observer’s
task, must be carefully tailored to achieve the experimental goal. A psychophysics experiment
consists of a number of components, and we have opted for the following breakdown: stim-
ulus; task; method; analysis; and measure (Figure 2.1). To illustrate our use of these terms,
consider one of the most basic experiments in the study of vision: the measurement of a
“contrast detection threshold.” A contrast detection threshold is defined as the minimum
amount of contrast necessary for a stimulus to be just detectable. Figure 2.2 illustrates the
idea for a stimulus consisting of a patch on a uniform background. The precise form of the
stimulus must, of course, be tailored to the specific question about sensory function being
asked, so we assume that the patch is the appropriate stimulus. The contrast of the patch
can be measured in terms of Weber contrast, defined as the difference between the luminance
of the patch and its background, DL, divided by the luminance of the background Lb, i.e., DL/
Lb. The contrast detection threshold is therefore the smallest value of Weber contrast needed
to detect the patch. Many procedures exist for measuring a contrast detection threshold, each
involving a different task for the observer. Before the advent of digital computers, a common

Psychophysics
experiment

Stimulus Task Method Analysis Measure

FIGURE 2.1 Components of a psychophysics experiment.
 2.2 TASKS, METHODS, AND MEASURES 13
1.00

0.90

Prop. correct
0.80

0.70

0.60

0.50
CT
0.40
ΔL Lb 0.00 0.05 0.10 0.15
Contrast

C=ΔL/Lb
FIGURE 2.2 Top left: circular test patch on a uniform background. Bottom left: luminance profile of the patch and
the definition of Weber contrast. Right: results of a standard two-interval-forced-choice (2IFC) experiment. The
various stimulus contrasts are illustrated on the abscissa. Black circles are the proportion of correct responses for
each contrast. The green curve is the best fit of a psychometric function, and the calculated contrast detection
threshold (CT) is indicated by the arrow. See text for further details. L ¼ luminance; Lb ¼ luminance of background;
DL ¼ difference in luminance between patch and background; C ¼ Weber contrast.

method was to display the stimulus on an oscilloscope and ask observers to adjust the
contrast with a dial until the stimulus was just visible. The just-visible contrast would then
be recorded as the contrast detection threshold. This method is typically termed the “method
of adjustment”, or MOA.
Nowadays the preferred approach is to present stimuli on a computer display and use a
“two-interval forced-choice,” or 2IFC, task. Using this procedure, two stimuli are presented
briefly on each trial, one of which is a blank screen, the other the test patch. The order of stim-
ulus presentationdblank screen followed by test patch or test patch followed by blank
screendis unknown to the observer (although of course “known” to the computer) and is
typically random or quasi-random. The two stimuli are presented consecutively, and the
observer chooses the interval containing the test patch, indicating his or her choice by press-
ing a key. The computer keeps a record of the contrast of the patch for each trial, along with
the observer’s response, which is scored as either “correct” or “incorrect.” A given experi-
mental session might consist of, say, 100 trials, and a number of different patch contrasts
would be presented in random order.
With the standard 2IFC task, different methods are available for selecting the con-
trasts presented on each trial. On the one hand, they can be preselected before the
experimentdfor example, 10 contrasts ranging from 0.01 to 0.1 at 0.01 intervals. If prese-
lected in this way, the 10 stimuli at each contrast would be presented in random order
during the session, making 100 trials in total. This is known as the “method of constants.”
At the end of each session the computer calculates the number of correct responses for
each contrast. Typically, there would be a number of sessions and the overall proportion
correct across sessions for each patch contrast calculated, then plotted on a graph as
shown for the hypothetical data in Figure 2.2. On the other hand, one could use an “adap-
tive” (or “staircase”) method, in which the contrast selected on each trial is determined by
14 2. CLASSIFYING PSYCHOPHYSICAL EXPERIMENTS

the observer’s responses on previous trials. The idea behind the adaptive method is that
the computer “homes in” on the contrasts that are close to the observer’s contrast detec-
tion threshold, thus not wasting too many trials on stimuli that are either too easy or too
hard to see. Adaptive methods are the subject of Chapter 5.
The term “analysis” refers to how the data collected during an experiment are converted
into measures. For example, with the method of adjustment the observer’s settings might be
averaged to obtain the threshold. On the other hand, using the 2IFC procedure in conjunction
with the method of constants, the proportion correct data may be fitted with a function whose
shape is chosen to match the data. The fitting procedure can be used to estimate the contrast
detection threshold defined as the proportion correct, say 0.75 or 75%, as shown in Figure 2.2.
Procedures for fitting psychometric functions are discussed in Chapter 4.
To summarize, using the example of an experiment aimed at measuring a contrast detection
threshold for a patch on a uniform background, the components of a psychophysical experi-
ment are as follows. The “stimulus” is a uniform patch of given spatial dimensions and of
various contrasts. Example “tasks” include adjustment and 2IFC. For the adjustment task, the
“method” is the method of adjustment, while for the 2IFC task one could employ the method
of constants or an adaptive method. In the case of the method of adjustment, the “analysis”
might consist of averaging the set of adjustments, whereas for the 2IFC task it might consist
of fitting a psychometric function to the proportion correct responses as a function of contrast.
For the 2IFC task in conjunction with an adaptive method, the analysis might involve averaging
contrasts, or it might involve fitting a psychometric function. The “measure” in all cases is a
contrast detection threshold, although other measures may also be extracted, such as an esti-
mate of the variability or “error” on the threshold and the slope of the psychometric function.
The term “procedure” is used ubiquitously in psychophysics and can refer variously to the
task, method, analysis, or some combination thereof. Similarly, the term “method” has broad
usage. The other terms in our component breakdown are also often used interchangeably.
For example, the task in the contrast detection threshold experiment, whether adjustment
or 2IFC, is sometimes termed a “detection” task and sometimes a “threshold” task, while
in our taxonomy the terms “detection threshold” refer to the output measure. The lesson
here is that one needs to be flexible in the use of psychophysics terminology and not overly
constrained by any predefined scheme.
Next we consider some of the common dichotomies used to characterize different psycho-
physical procedures and experiments. The aim here is to introduce some common terminol-
ogy, illustrate other varieties of psychophysical experiment besides contrast detection, and to
examine which, if any, of the dichotomies might be candidates for a psychophysics classifi-
cation scheme.

2.3 DICHOTOMIES

2.3.1 “Class A” versus “Class B” Observations

An influential dichotomy introduced some years ago by Brindley (1970) is that between
“Class A” and “Class B” psychophysical observations. Although one rarely hears these terms
today, they are important to our understanding of the relationship between psychophysical
measurement and sensory function. Brindley used the term “observation” to describe the
 2.3 DICHOTOMIES 15
Class A

Adjust Adjust

Class B

Adjust

FIGURE 2.3 The Rayleigh match illustrates the difference between a Class A and Class B psychophysical
observation. For Class A, the observer adjusts both the intensity of the yellow light in the right half of the bipartite
field as well as the relative intensities of the red and green lights in the mixture in the left half of the bipartite field
until the two halves appear identical. For Class B, the observer adjusts only the relative intensities of the red and
green lights in the left half to match the hue of a yellow light in the right half that in this example is different in
brightness.

perceptual state of an observer while executing a psychophysical task. The distinction be-
tween Class A and Class B attempted to identify how directly a psychophysical observation
related to the underlying mental processes. Brindley framed the distinction in terms of a com-
parison of sensations: a Class A observation refers to the situation in which two physically
different stimuli are perceptually indistinguishable, whereas a Class B observation refers to
all other situations.
The best way to understand the difference between Class A and Class B is with an
example, and for this we have adopted Gescheider’s (1997) example of the Rayleigh match
(Rayleigh, 1881; Thomas and Mollon, 2004). Rayleigh matches are used to identify and study
certain types of color vision deficiency (e.g., Shevell et al., 2008), but for the present purposes
the aim of a Rayleigh match is less important than the nature of the measurement itself.
Figure 2.3 shows a bipartite circular stimulus, one half consisting of a mixture of red and
green monochromatic lights, the other half a yellow monochromatic light.1 During the
1
Because the lights are monochromatic, i.e., narrow band in wavelength, this experiment cannot be
conducted on a CRT (cathode ray tube) monitor, because CRT phosphors are relatively broadband in
wavelength. Instead an apparatus is required that can produce monochromatic lights, such as a Nagel
Anomaloscope or a Maxwellian view system.
16 2. CLASSIFYING PSYCHOPHYSICAL EXPERIMENTS

measurement procedure the observer is given free reign to adjust both the intensity of the yel-
low light as well as the relative intensities of the red and green lights. The task is to adjust the
lights until the two halves of the stimulus appear identical, as illustrated in the top of the
figure. In color vision, two stimuli with different spectral (i.e., wavelength) compositions
but that appear identical are termed “metamers.” According to Brindley, metameric matches
such as the Rayleigh match are Class A observations. The identification of an observation as
Class A accords with the idea that when two stimuli appear identical to the eye they elicit
identical neural responses in the brain. Since the neural responses are identical, Brindley
argues, it is relatively straightforward to map the physical characteristics of the stimuli
onto their internal neural representations.
An example of a Class B observation is shown at the bottom of Figure 2.3. This time the
observer has no control over the intensity of the yellow light, only control over the relative
intensities of the red and green lights. The task is to match the hue (or perceived chromaticity)
of the two halves of the stimulus but with the constraint that the intensity (or brightness) of the
two halves remains different. Thus, the two halves will never appear identical and therefore,
according to Brindley, neither will the neural responses they elicit. Brindley was keen to point
out that one must not conclude that Class B observations are inferior to Class A observations:
our example Class B observation is not a necessary evil due to defective equipment! On the con-
trary, we may wish to determine the spectral combinations that produce hue matches for stim-
uli that differ in brightness, precisely to understand how hue and brightness interact in the
brain. In any case, the aim here is not to judge the relative merits of Class A and Class B obser-
vations (for a discussion of this see Brindley, 1970) but rather to illustrate what the terms mean.
What other types of psychophysical experiment are Class A and Class B? According to
Brindley, experiments that measure thresholds, such as the contrast detection threshold
experiment discussed in the previous section, are Class A. This might not be intuitively
obvious, but the argument goes something like this. There are two states: stimulus present
and stimulus absent. As the stimulus contrast is decreased to a point where it is below
threshold, the observation passes from one in which the two states are discriminable to
one in which they are indiscriminable. The fact that the two states may not be discriminable
even though they are physically different (the stimulus is still present even though below
threshold) makes the observation Class A. Two other examples of Class A observations
that accord to the same criterion are shown in Figure 2.4.
Class B observations characterize many types of psychophysical procedure. Following
our example Class B observation in Figure 2.3, any experiment that involves matching
two stimuli that are perceptibly different on completion of the match is Class B. Consider,
for example, the brightness-matching experiment illustrated in Figure 2.5. The aim of this
experiment is to determine how the brightness, i.e., perceived luminance, of a test disk is
influenced by the luminance of its surround. As a rule, increasing the luminance of a sur-
round annulus causes the disk inside to decrease in brightness, i.e., become dimmer. One
way to measure the amount of dimming is to adjust the luminance of a second, matching
disk until it appears equal in brightness to the test disk. The matching disk can be thought
of as a psychophysical “ruler.” When the matching disk is set to be equal in brightness to
the test disk, the two disks are said to be at the “point of subjective equality,” or PSE. The
luminances of the test and match disks at the PSE will not necessarily be the same; indeed it
is precisely because they are as a rule different that is of interest. The difference in
 2.3 DICHOTOMIES 17

FIGURE 2.4 Two other examples of Class A observations. Top: orientation discrimination task. The observer is
required to discriminate between two gratings that differ in orientation, and a threshold orientation difference is
measured. Bottom: line bisection task. The observer is required to position the vertical red line midway along the
horizontal black line. The precision or variability in the observer’s settings is a measure of his or her line-bisection
acuity.

luminance between the test and match disks at the PSE tells us something about the effect
of context on brightness, the “context” in this example being the annulus. This type of
experiment is sometimes referred to as “asymmetric brightness matching,” because the
test and match disks are situated in different contexts (e.g., Blakeslee and McCourt,
1997; Hong and Shevell, 2004).
It might be tempting to think of an asymmetric brightness match as a Class A observa-
tion, on the grounds that it is quite different from the Class B version of the Rayleigh match
described above. In the Class B version of the Rayleigh match, the stimulus region that

(a) Match Test

(c)

(b)

FIGURE 2.5 Two examples of Class B observations. In (a) the goal of the experiment is to find the point of
subjective equality (PSE) in brightness between the fixed test and variable match patch as a function of the lumi-
nance (and hence contrast) of the surround annulus; (b) shows the approximate luminance profile of the stimulus;
(c) is the MullereLyer illusion. The two center lines are physically identical but appear different in length. The
experiment described in the text measures the relative lengths of the two vertical axes at which they appear equal in
length.
18 2. CLASSIFYING PSYCHOPHYSICAL EXPERIMENTS

observers match in hue is also the region that differs along the other dimensiondbrightness.
In an asymmetric brightness-matching experiment on the other hand, the stimulus region
that observers match, brightness, is not the region that differs between the test and match
stimuli - in this instance it is the annulus. However, one cannot “ignore” the annulus when
deciding whether the observation is Class A or Class B simply because it is not the part of
the stimulus to which the observation is directed. Asymmetric brightness matches are Class
B because, even when the stimuli are matched, they are recognizably different by virtue of
the fact that one stimulus has an annulus and the other does not.
Another example of a Class B observation is the MullereLyer illusion shown in
Figure 2.5(c), a geometric illusion that has received considerable attention (e.g., Morgan
et al., 1990). The lengths of the axes in the two figures are the same, yet they appear different
due to the arrangement of the fins at either end. One of the methods for measuring the size
of the illusion is to require observers to adjust the length of the axis, say of the fins-inward
stimulus, until it matches the perceived length of the axis of the other, say fins-outward stim-
ulus. The physical difference in length at the PSE, which could be expressed as a raw, propor-
tional, or percentage difference, is a measure of the size of the illusion. The misperception of
relative line length in the MullereLyer figures is a Class B observation, because even when
the lengths of the axes are adjusted to make them perceptually equal, the figures remain
perceptibly different as a result of their different fin arrangements.
Another example of a Class B observation is magnitude estimation. This is the procedure
whereby observers provide a numerical estimate of the perceived magnitude of a stimulus,
for example along the dimension of contrast, speed, depth, size, etc. Magnitude estimation
is Class B because our perception of the stimulus and our judgment of its magnitude utilize
different mental modalities.
An interesting case that at first defies classification into Class A or Class B is illustrated
in Figure 2.6. The observer’s task is to discriminate the mean orientation of two random
arrays of line elements, whose mean orientations are right- and left-of-vertical (e.g., Dakin,
2001). Below threshold, the mean orientations of the two arrays are indiscriminable, yet
the two arrays are still perceptibly different by virtue of their different element arrangements.
In the previously mentioned Class B examples, the “other” dimensiondbrightness in the
case of the Rayleigh match, annulus luminance in the case of the brightness-matching
experimentdwas relevant to the task. However in the mean-orientation-discrimination
experiment the “other” dimensiondelement positiondis irrelevant. Does the fact that

FIGURE 2.6 Class A or Class B? The observer’s task is to decide which of the two stimuli contains elements that
are on average left-oblique. When the difference in mean element orientation is below threshold, the stimuli are
identical in terms of their perceived mean orientation, yet are discriminable on the basis of the arrangement of their
elements.
 2.3 DICHOTOMIES 19
element arrangement is irrelevant make it Class A, or does the fact that the stimuli are
discriminable below threshold on the basis of element arrangement make it Class B? Readers
can decide.
In summary, the Class A versus Class B distinction is important for understanding the
relationship between psychophysical measurement and sensory function. However, we
choose not to use this dichotomy as a basis for classifying psychophysics experiments, in
part because there are cases that seem hard to classify in terms of Class A or Class B, and
in part because other dichotomies for us better capture the critical differences between psy-
chophysical experiments.

2.3.2 “Type 1” versus “Type 2”

An important consideration in sensory measurement concerns whether or not an ob-
server’s responses can be designated as “correct” or “incorrect”. If they can be so designated,
the procedure is termed Type 1 and if not Type 2 (Sperling, 2008; see also Sperling et al.,
1990). The term Type 2 has sometimes been used to refer to an observer’s judgments about
their own Type 1 decisions (Galvin et al., 2003); in this case, the Type 2 judgment might be
a rating of, say, 1e5, or a binary judgment such as “confident” or “not confident,” in refer-
ence to their Type 1 decision2.
The forced-choice version of the contrast threshold experiment described earlier is a pro-
totypical Type 1 experiment, whereas the brightness-matching and MullereLyer illusion ex-
periments, irrespective of whether or not they employ a forced-choice procedure, are
prototypical Type 2 experiments. There is sometimes confusion, however, as to why some
forced-choice experiments are Type 2. Consider again the MullereLyer illusion experiment.
As with the contrast detection threshold experiment, there is more than one way to measure
the size of the illusion. We have already described the adjustment procedure. Consider how
the MullereLyer might be measured using a forced-choice procedure. One method would be
to present the two fin arrangements as a forced-choice pair on each trial, with the axis of one
fixed in length and the axis of the other variable in length. Observers would be required on
each trial to indicate the fin arrangement that appeared to have the longer axis. Figure 2.7
shows hypothetical results from such an experiment. Each data point represents the propor-
tion of times the variable-length axis is perceived as longer than the fixed-length axis, as a
function of the length of the latter. At a relative length of 1, meaning that the axes are phys-
ically the same, the observer perceives the variable axis as longer almost 100% of the time.
However, at a relative axis length of about 0.88, the observer chooses the variable axis as
longer only 50% of the time. Thus, the PSE is 0.88. However, even though the MullereLyer
experiment, like the contrast threshold experiment, can be measured using a forced-choice
procedure, there is an important difference between the two experiments. Whereas in the
contrast detection threshold experiment there is a correct and an incorrect response on every
trial, there is no correct or incorrect response for the MullereLyer trials. Whatever response
the observer makes on a MullereLyer trial, it is meaningless to score it as correct or incorrect,
at least given the goal of the experiment, which is to measure a PSE. Observers unused to
doing psychophysics often have difficulty grasping this idea and even when told repeatedly
2
Note that the dichotomy is not the same as Type I and Type II errors in statistical inference testing.
20 2. CLASSIFYING PSYCHOPHYSICAL EXPERIMENTS

1.00

Proportion longer responses

0.75

0.50

0.25

0.00
0.8 0.9 1.0 1.1 1.2
Variable/fixed length ratio

Fixed
Variable
FIGURE 2.7 Results of a hypothetical experiment aimed at measuring the size of the MullereLyer illusion using a
forced-choice procedure and the method of constant stimuli. The critical measurement is the PSE between the lengths
of the axes in the fixed test and variable comparison stimuli. The graph plots the proportion of times subjects perceive
the variable axis as “longer.” The continuous line through the data is the best-fitting logistic function (see Chapter 4).
The value of 1.0 on the abscissa indicates the point where the fixed and variable axes are physically equal in length.
The PSE is calculated as the variable axis length at which the fixed and variable axis lengths appear equal, indicated
by the vertical green arrow. The horizontal red-arrowed line is a measure of the size of the illusion.

that there are no correct and incorrect answers, insist on asking at the end of the experiment
how many trials they scored correct!
The Type 1 versus Type 2 dichotomy is not synonymous with Class A versus Class B,
though there is some overlap. For example, the Rayleigh match experiment described above
is Class A but Type 2 because no “correct” match exists. On the other hand, the two-alternative
forced-choice (2AFC) contrast threshold experiment is both Class A and Type I.
The Type 1 versus Type 2 dichotomy is an important one in psychophysics. It dictates, for
example, whether observers can be provided with feedback during an experiment, such as a
tone for an incorrect response. However, one should not conclude that Type 1 is “better” than
Type 2. The importance of Rayleigh matches (Class A but Type 2) for understanding color
deficiency is an obvious case in point.

2.3.3 “Performance” versus “Appearance”

A dichotomy related to Type 1 versus Type 2, but differing from it in important ways, is
that between “performance” and “appearance.” Performance-based tasks measure aptitude,
i.e., “how good” an observer is at a particular task. For example, suppose one measures
contrast detection thresholds for two sizes of patch, call them “small” and “big.” If thresh-
olds for the big patch are found to be lower than those for the small patch, one can conclude
that observers are better at detecting big patches than small ones. By the same token,
if orientation discrimination thresholds are found to be lower in central than in peripheral
 2.3 DICHOTOMIES 21
vision, one can conclude that orientation discrimination is better in central vision than in the
periphery. Both of the above tasks aim to establish the limits of our perception. On the other
hand, suppose we measure the size of the MullereLyer illusion for two different fin angles,
say 45 and 60 (relative to the axis), and find that the illusion is bigger for the 45 fins. It
would be meaningless to conclude that we are “better” at the MullereLyer task when it has
45 compared to 60 fins. PSEs are not aptitudes. For this reason the MullereLyer experi-
ment is best considered as measuring stimulus appearance. A simple heuristic can be
used to decide whether a psychophysical procedure measures performance or appearance.
If the end measurement can be meaningfully considered as showing greater aptitude for
one condition than another, then it is measuring performance, and if not, appearance.
This still leaves open the question of a precise definition of appearance, other than “not per-
formance.” The term appearance, however, is not easy to define, but for most of the situa-
tions described in this book appearance can be defined as the apparent magnitude of a
stimulus dimension.
Sometimes the same psychophysical procedure can be used to measure both performance
and appearance. Consider the Vernier alignment task illustrated in Figure 2.8, applied to
two stimulus arrangements, labelled A and B. The goal of the experiment using stimulus
A is to measure Vernier acuity, which is defined as the smallest misalignment that can be
detected. This is a threshold measure and hence a performance measure. The goal of the
experiment using stimulus B is to measure the effect of the flanking white lines on the
perceived position of the black lines. The white lines in B tend to have a small repulsive ef-
fect, causing the black lines to appear slightly shifted from their normal perceived position,
in a direction away from that of the white lines (e.g., Badcock and Westheimer, 1985). For
both experiments, however, the task is the same: decide on each trial whether the upper
black line lies to the left (or to the right) of the lower black line.

FIGURE 2.8 Left: stimulus arrangements A and B for two Vernier alignment experiments. Right: hypothetical
data from each experiment. The abscissa plots the horizontal physical separation between the black lines, with
positive values indicating that the top line is physically to the left of the bottom line and negative values indicating
that the top line is physically to its right. The ordinate gives the proportion of times the observer responds that the top
line is “left.” The continuous curves are best-fitting logistic functions. The green arrow indicates for stimulus A the
Vernier threshold and the red arrow indicates for stimulus B the point-of-subjective alignment.
22 2. CLASSIFYING PSYCHOPHYSICAL EXPERIMENTS

Hypothetical data for A and B are shown in the graph on the right. The data points have
been fitted with logistic functions, as described in Chapter 4. For the A data, Vernier acuity
can be calculated as the line separation producing a proportion of 0.75 “left” responses, indi-
cated by the green arrow. Sometimes, however, the observer will perceive the two lines as
aligned even when they are physically misaligned. In other words, the point-of-subjective-
alignment, or PSA will not be zero. A nonzero PSA may result from optical aberration in
the eye or because the observer’s internal representation of space is nonveridical or because
the monitor display is physically distorted. It therefore makes more sense to measure Vernier
acuity as the separation (or half the separation) between the points on the abscissa corre-
sponding to the 0.25 and 0.75 response levels, as this takes into account any nonzero PSA.
Alternatively, the measure of Vernier acuity can be the steepness, or slope, of the psychomet-
ric function. As mentioned earlier, the slope of the psychometric function is inversely related
to the standard deviation of the function used to fit the data, so the standard deviation is
therefore also a measure of (the inverse of) Vernier acuity (e.g., Watt and Morgan, 1983;
McGraw et al., 2004). Recall also that the standard deviation is a measure of precision,
with a smaller standard deviation indicating a higher precision. Whether the threshold or
slope is used as the measure of Vernier acuity, however, both are performance measures
since the “better than” heuristic applies. Note, however, that because the PSA might be
nonzero, it is best not to regard the experiment using stimulus A as Type 1, i.e., as having
a correct and an incorrect response on each trial. Why? Suppose that when physically
aligned, an observer perceives the upper line as slightly to the left of the lower line. On trials
where the upper line is presented slightly to the right, the observer will tend to respond
“left” and if the experiment is treated as Type I, scored “incorrect.” If correct-versus-
incorrect feedback is provided to the observer this will inevitably cause confusiondafter
all, the observer really did see those lines as “left”dand the confusion could be detrimental
to performance.
The fact that a performance measure such as Vernier acuity is best measured without
feedback exemplifies how the distinction between performance and appearance is not
synonymous with Type 1 and Type 2. Moreover, precision, which we have argued is
a performance measure, can be obtained from any Type 2 experiment measuring a
PSE. Other examples of performance measures not necessarily derived from Type 1 ex-
periments are contrast detection thresholds obtained using the method of adjustment,
measures of accuracy (see next paragraph), and measures of reaction time. Thus,
although all Type 1 experiments measure performance, not all performance measures
are obtained from Type 1 experiments. On the other hand, all experiments that measure
appearance are Type 2.
Not only the precision but also the bias in the Vernier alignment experiment using stim-
ulus A can be considered as a measure of performance. The bias is measured in relation to
the true physical alignment, and so one can define the accuracy of the measure as its closeness
to the true alignment. Accuracy is important to vision, for example when estimating distances
and other spatial relationships as one navigates the visual world. For the Vernier experiment,
the bigger the bias the lower the accuracy. A similar argument holds for the line bisection task
illustrated in Figure 2.4. In this case, accuracy is how close the observer’s mean setting is to
the physical midpoint of the line. Since one can legitimately argue that one observer is more
accurate than another in either Vernier alignment or line bisection, the accuracy measured in
 2.3 DICHOTOMIES 23
these tasks is a performance measure. However, as we shall now see, measures of bias in
many circumstances are better considered to be measures of appearance.
Consider the Vernier alignment task using stimulus B. As with the Mullere Lyer and
brightness-matching experiments, it is the bias that we are primarily interested in. We
want to know by how much the PSA is shifted by the presence of the white lines. The shift
in the PSA is measured as the separation between the PSAs for stimuli A and B, with each
PSA calculated as the point on the abscissa corresponding to 50% “left” responses.
Assuming that the PSA with stimulus A is at zero, the shift in PSA caused by the white lines
is indicated by the green arrow on the graph associated with stimulus B. This shift is a mea-
sure of appearance.
Innumerable aspects of stimulus appearance avail themselves to psychophysical measure-
ment, for example choosing the computer sketch of a stimulus that best matches its appear-
ance (e.g., Georgeson, 1992); indicating when a simulated three-dimensional random-dot
rotating cylinder appears to reverse direction (e.g., Li and Kingdom, 1999); adjusting the
colors of a moving chromatic grating until the grating appears to almost stop (Cavanagh
et al., 1984); and labeling contour-defined regions in images of natural scenes as being either
“figure” or “ground” (e.g., Fowlkes et al., 2007). Are there any broad classes of procedure that
measure appearance? Matching and scaling experiments are arguably example classes.
Matching experiments measure PSEs between two physically different stimuli, as in the Ray-
leigh match, brightness-matching, MullereLyer, and Vernier task B experiments described
above. Scaling experiments, the topic of Chapter 8, determine the relationship between the
perceived and physical dimensions of a stimulus. Example perceptual scales are the relations
between perceived and physical contrast, hue (or perceived chromaticity) and wavelength,
perceived and physical velocity, and perceived depth and retinal disparity. Although not
all perceptual scales are appearance-based, most of them are.
Example data from a scaling experiment are shown in Figure 2.9. Unlike the hypothetical
data used so far to illustrate generic experimental results, every perceptual scale has a unique
shape, so for Figure 2.9 we have reproduced a specific case from an experiment conducted by
Whittle (1992). Whittle was interested in the relationship between the brightness (or
perceived luminance) and the physical luminance of discs on a gray background. Observers
were presented with a display consisting of 25 discs arranged in a spiral, with the first and
last fixed in luminance at respectively the lowest and highest available on the monitor, cor-
responding to “black” and “white.” Observers adjusted the luminances of the remaining 23
discs until they appeared to be at equal intervals in brightness. Figure 2.9 plots the disc num-
ber (1e25) against the resulting luminance settings. If brightness (the perceptual dimension)
was linearly related to luminance (the physical dimension) then the function would be a
straight line. Instead it has a complex shape. The reason for this particular shape is another
story (see Kingdom and Whittle, 1996); our aim here is merely to illustrate a type of percep-
tual scale. There are many different procedures for deriving perceptual scales, and these are
summarized in Chapter 3, with further details in Chapter 8.
Both performance-based and appearance-based experiments are important to our un-
derstanding of vision. Measures from both types of experiment are necessary to charac-
terize the system. The relationship between performance and appearance, and the
question as to what each tells us about visual function, is an important but complex issue
that is beyond the remit of this book (e.g., in some instances they appear to measure
24 2. CLASSIFYING PSYCHOPHYSICAL EXPERIMENTS

Number of disc

Luminance cd/m2
FIGURE 2.9 Data from a brightness scaling experiment. The graph plots the number of the disc against its
luminance, after the luminances of all the discs have been adjusted to make them appear at equal brightness intervals.
The green arrow indicates the point where the discs change from being decrements to increments. Data based on
Whittle (1992).

closely related sensory processes, such as the luminance-discrimination threshold and

brightness scaling results compared in Whittle (1992), while in other instances they deal
with different processes, as argued by Gheorghiu and Kingdom (2008) in relation to cur-
vature perception). However, we argue that the performance versus appearance dichot-
omy more than any other dichotomy is the principle dividing line in psychophysics.
For this reason we propose it as the candidate for the superordinate division in our clas-
sification scheme. In the next section, we discuss a possible second level of categorization
in the scheme.

2.3.4 “Forced-Choice” versus “Nonforced-Choice”

By now the reader should be familiar with the concept of the forced-choice procedure, but
as with many of the terms in psychophysics, the devil lies in the details. In particular, there
are different conventions as to when one should and when one should not use the term
“forced-choice” and different conventions for the number of alternatives/intervals that prefix
the term. In Signal Detection Theory (Wickens, 2002; McNicol, 2004; Macmillan and Creel-
man, 2005), the subject of Chapters 6 and 7, “forced-choice” is mainly used to characterize
experiments in which two or more stimulus alternatives are presented during a trial, one
of which is the “target.” Example forced-choice tasks that accord with this usage are: deciding
which of two stimuli, a blank field or a patch, contains the patch; deciding which of two
patches is brighter; and deciding which of three lines, two oriented 5 and one
oriented þ5 , is the 5 line. In these examples, the observer is required to select a stimulus
from two or more stimuli during each trial. Typically, at the end of the experiment the pro-
portion of trials in which the target alternative was selected is calculated for each stimulus
Other documents randomly have
different content
He was gone until nearly ten o’clock, and the time dragged heavily to
Neva’s friends, who remained in their closed sitting-room,
exchanging surmises and doubts, and preparing themselves for an
encounter with Craven Black and Octavia.
Sir Harold put on his greatcoat and turned up his collar, and wound a
gray woolen muffler about the lower part of his face. He was
standing thus disguised, hat in hand, when Ryan came back and
quietly slipped into the room.
“The cab is waiting,” announced the detective. “I have been at Rufus
Black’s heels ever since I left you. When I got back to the
Caledonian, he was just going out in his cab. I rode on top as a
friend of the driver, who was won over to make a friend of me by a
gift of a crown. We drove to the minister’s, and to the sexton’s, and
finally to a jeweller’s, where Black bought a ring. We then went back
to the hotel. And a few minutes ago young Black entered his cab
again, and gave the order ‘to the church.’ I know the church, and we
must get on our way to reach it, if we expect to get there in time to
stop the ceremony.”
Sir Harold and Lord Towyn hurried impetuously out of the room and
down the stairs, and were seated in the cab when Atkins and the
detective reached the street. These two also entered the vehicle,
which rolled swiftly down the street.
A few minutes’ drive brought them to the plain, substantial kirk which
had been chosen by Rufus Black as the scene of his second
marriage to Lally.
The four pursuers leaped from the cab, and hastily entered the
edifice by its half-opened door.
Passing through the dim and chilly vestibule, they pushed open one
of the baize-covered inner doors, which swung noiselessly upon its
well-oiled hinges, and stood within the kirk.
It was a plain church interior, without stained glass or lofty arched
window, with bare walls and ceilings, and with the plainest of
gasoliers; the reading desk of solid oak, beautifully carved, was yet
in keeping with the rugged simplicity of this house of worship. Here
the old Covenanters might have worshiped; and here their
descendants did worship, in all the stern simplicity of the faith in
which they had been trained.
There was no one save the pew-opener in the church at the moment
of the intrusion of Sir Harold Wynde and his companions. The four
passed silently down the long dim aisle, and entered a tall-backed
pew, in which they were nearly hidden from view. Lord Towyn gave
the pew-opener a shilling, and they were left to themselves.
“It doesn’t look like a wedding,” said Sir Harold, shivering in his
greatcoat. “If the bridegroom came on before us, where is he?”
The question was answered by the appearance of Rufus Black and
the minister from the little vestry, in which they had gone to warm
themselves. Rufus wore his ordinary garments, but had bought a
white waistcoat and neck-tie, which gave him a clerical air. He kept
his eyes upon the door with an anxious, uneasy glance.
“He’s afraid she’ll give him the slip, after all,” muttered Ryan.
The green baize-covered door swung open and closed again. Rufus
Black and the occupants of the high-back pew in the corner near the
reading desk alike started, but the arrival was only that of a few
persons who had seen the open church door, and surmised that a
wedding was in progress. They questioned the pew-opener, and
subsided into pews.
Presently a few more curious persons appeared, and took their seats
also.
The occupants of the high-backed pew grew impatient. It was after
ten o’clock, an early hour for a wedding, but the hour Rufus had
himself appointed, in his eager impatience to claim his young wife. A
cold sweat started to the young bridegroom’s face. He began to think
that Lally had thought better of her promise to remarry him, and had
decided to give him up for the worthless, weak, irresolute being he
knew himself to be.
“There’s a hitch somewhere,” said Ryan.
Again the baize-covered door swung open, and four persons came
slowly up the aisle.
It was the bridal party at last.
Rufus Black started forward with an irrepressible eagerness, joy and
relief. Sir Harold Wynde and Lord Towyn, alike pale and agitated,
regarded the approaching party with burning eyes.
First of all came the steward of Heather Hills, with a girlish figure
clinging to his arm. Behind these two came the steward’s wife in gray
silk, and Peters in black silk and crape, but with white ribbons at her
throat, and white lace collar and sleeves.
Sir Harold and the young earl looked at the three strange figures in a
sort of bewilderment. They had expected to see Craven Black and
Octavia. Not seeing them, they fixed their glances upon Lally.
The young wife had laid aside her mourning for her great-aunt upon
this occasion, and wore a dress that Mrs. Wroat had bought for her
upon their memorable shopping expedition immediately after Lally’s
arrival in London.
It was a delicate mauve moire, made with a long train. Over it was
worn an upper dress of filmy tulle, arranged in foam-like puffs over
all its surface. This too formed a trail. The corsage was of puffs of
tulle over the moire, and was made low in the neck and short in the
sleeves. The bride wore a tulle vail, which fell over her face in soft
folds, and was confined to her head by an aigrette of diamonds.
Through the filmy folds of her vail the spectators caught the gleam of
diamonds on her arms and neck and bosom.
The steward conducted his beautiful young charge to the altar, and
bride and bridegroom stood side by side and the minister slowly took
his place.
Lord Towyn made a movement to dash from his seat, but Sir Harold
caught his arm in a stern grip, and compelled him to remain.
At the moment of beholding the bride, a mist had swept over the
young earl’s vision. His brain had seemed to swim. For the instant he
had scarcely doubted, in his excitement, that it was Neva who stood
before him; but as his vision cleared, he knew that this young bride
was not his betrothed wife. He knew it, although he could not see
Lally’s face. He missed the haughty carriage of Neva’s slender
figure, the proud poise of her small, noble head, the swaying grace
of her movements. This young bride was not so tall as Neva, and
had not Neva’s dainty imperial grace.
“It’s not Neva!” he whispered excitedly. “That is Rufus Black, sure
enough, but the lady is not Neva.”
“You can’t see her face,” said Atkins. “I think it is Miss Wynde.”
At this moment the bride with a sweep of her hand, threw back her
vail. As her bright, dark face, so like a gipsy’s and with a glow of
happiness upon it, met the gaze of the spectators, Sir Harold stifled
a groan.
Lord Towyn stared at the pretty brown face, with its fluctuating color,
and the softly melting black eyes, and a dead pallor covered his
face.
If this young girl was the chosen bride of Rufus Black, where was
Neva? Why had Rufus given her up? The wildest fears for her life
and safety possessed him.
The marriage went on. The four pursuers who had come to interrupt
the proceedings sat in their high-backed pew as if utterly stupefied.
What objection could they raise to the marriage of Rufus Black to a
stranger who came to the church escorted by her friends? Why
should they object to such a marriage? They heard the questions
and answers as in a trance. The name of Lalla Bird sounded
strangely upon their ears. And when the minister said, “I now
pronounce you man and wife, and whom God has joined together let
not man put asunder,” Sir Harold Wynde and the young earl looked
at each other with terrified, inquiring eyes, that asked the question
that filled their souls alike: Where was Neva?
After the prayer that followed the ceremony, the minister went into
the vestry, followed by the newly married pair, the steward and his
wife, and good Mrs. Peters.
The casual spectators of the wedding stole silently out of the church.
“Well, I’ve come up here on a fool’s errand,” muttered Ryan, in a
tone of chagrin.
“Perhaps not,” said Lord Towyn. “Rufus may be able to give us some
clue to his father’s whereabouts, if we approach him judiciously. I am
going into the vestry to see him.”
“And I too,” said the baronet, rising.
The young earl led the way from their pew to the vestry, Sir Harold at
his side, and Mr. Atkins and Ryan behind them.
The bride, all blushes and smiles, was writing her name in the
marriage register, when the young earl and his companions entered
the small room. Rufus Black had just signed it, and was putting on
his gloves. He gave a great start as he recognized Lord Towyn and
Atkins, and stared beyond them with an unmistakable terror, as if he
expected to behold the cynical sneering face and angry eyes of his
father looming up behind the intruders.
“You here, my lord?” he faltered.
“Yes, Rufus,” said the young earl, holding out his hand. “We
happened to be at Inverness and have been witnesses to your
marriage. Permit us to congratulate you.”
Rufus drew a long breath of relief and shook the earl’s hand heartily.
“I thought—I thought—” he began, confused and hesitating—“I was
afraid— But never mind. It’s odd your being up here, my lord. How
do you do, Atkins? Lally,” and Rufus turned to his young wife, who
was looking curiously at the new-comers, “here are some friends of
mine, up from Kent. Lord Towyn, Lally, and Mr. Atkins of Canterbury.”
Lally blushed and acknowledged the introduction gracefully.
“Can we see you in the church a moment, Mr. Black?” asked Lord
Towyn.
Rufus consented, with that look of fear again in his eyes. He
apologized for a moment’s absence to his bride and her friends, who
were now signing their names to the register, and accompanied his
pursuers back into the church. His face brightened when he found
that his father was not in waiting for him in the church.
“You have a pretty young bride, Rufus,” said the young earl
pleasantly. “We have followed you up from Kent, with the idea that
you were on your way to Miss Wynde. It was in this way we
happened to be at your wedding. Is Mrs. Rufus Black a recent
acquaintance?”
Rufus hesitated, with a quick glance at Sir Harold’s muffled face and
figure. Then he said bravely, resolving to act upon his new principles
of straightforwardness and courage:
“It is an odd story, Lord Towyn. I have been married before to my
wife to whom I was married this morning. My father separated me
from her and I read in a London paper that she was dead. I
discovered my mistake the other day in London. I met her in a
picture-shop. She came off to Scotland that night, and I found her
yesterday. She is an heiress now, my lord, but the same true and
loving wife she used to be. I was desperate at her loss; I was half
mad, I think, when I asked Miss Wynde to marry me. I never loved
any one but my own wife, and I beg you to say to Miss Wynde for
me, that I send my best wishes for her happiness, and I should be
glad to witness her marriage with you, my lord.”
“Thank you, Rufus. But where is Miss Wynde?”
A look of genuine surprise appeared in Rufus Black’s eyes.
“Why, she is at Wynde Heights, with my father and her step-mother,”
he answered.
“She is not there. They have not been there. They have conveyed
her to some lonely place, where they hope to subdue her into
consenting to marry you,” said Lord Towyn. “Can you give us no clue
to their whereabouts?”
“None whatever, my lord. My father said they were going to Wynde
Heights, and ordered me to hold myself in readiness to come to him
at a moment’s warning. I have not heard from him since he left
Hawkhurst. I am now of age, and have flung off my father’s authority
forever. I know no more than you do, my lord, where my father can
have gone. But one thing is sure. When he sees the announcement
of my marriage in the Times, he’ll give up the game, and bring Miss
Wynde back to her home.”
“He may not dare to do that,” said Atkins. “He has carried matters
with too high a hand, and has gone too far to make an easy retreat.
Has your father any property, Mr. Black?”
“About three hundred a year,” said Rufus. “His wife is rich.”
“I mean, does he own any real estate?”
Rufus smiled, shook his head, and hesitated.
“I don’t know,” he said doubtfully. “I believe he does own a small
estate somewhere, but it never brought him in a penny. It is barren,
unproductive, and out of the world.”
“The very place to which he would have gone!” cried Atkins. “Where
is it?”
“I don’t know,” confessed Rufus. “You see my father never talked of
his affairs to me. In fact, I never lived with him. I was always at
school, and we were more like strangers, or master and serf, than
like father and son. His property may be in Wales, and it may be
elsewhere. I believe there are mountains near it or around it, but I
am not sure. Indeed, my lord, almost any one who ever knew my
father can tell you more about his affairs than I can.”
Rufus spoke with a plain sincerity that convinced his hearers of his
truthfulness.
“We have had our journey to Scotland for nothing,” said Atkins.
Ryan looked crestfallen.
“We will detain you no longer, Rufus,” said Lord Towyn, a shadow
darkening his fair and noble face. “Make our excuses to your bride
for taking you from her so soon after your marriage, and accept our
best wishes for your future prosperity and happiness. And now good-
morning.”
With an exchange of courtesies and friendly greetings the party
broke up, Rufus Black returning to his young wife and her wondering
friends, to make all necessary apologies for his absence from them,
and Lord Towyn and his companions making their way into the
street.
“What are we to do now?” demanded the young earl, as they paused
at the open door of the cab.
Sir Harold looked at his daughter’s lover with haggard eyes.
“I am worn out with excitement and fatigue,” said the baronet, in a
low, weary voice. “I will go back to the hotel and lie down. I must not
become worn out. Heaven knows I shall need all my strength.”
“And you, Atkins?” said Lord Towyn.
“I shall try to catch a nap also,” said the solicitor gloomily. “I’m tired
too. I can’t stand it to go banging back to Kent by the first south-
bound train.”
“And you, Ryan?” asked the earl.
“I don’t know,” said the detective. “I want to think over what has
happened, and see if I can get any new ideas.”
He raised his hat, and walked away.
“I’ll take a stroll about the town, Sir Harold,” said the earl. “I feel
strangely restless, and not at all sleepy. I slept very well last night on
the train—as well as I have slept since Neva disappeared. I’ll meet
you and Atkins in our sitting-room at the Railway Hotel by four
o’clock.”
Sir Harold and Atkins entered the cab, and were driven to their hotel.
The young earl watched the cab until it disappeared from sight, and
then he walked down the street, idly taking his way toward the river.
The wind blew strong and fiercely—a very winter wind, as cold and
keen as if it blew directly from the North pole, and having
suggestions of icebergs in it. The young earl shivered, and drew up
his coat collar.
“Pretty weather for this season,” he muttered. “The gale of night
before last has not quite blown itself out, and is giving us a few
parting puffs.”
He walked down to the wharves and stood by the water’s edge, his
hat pulled over his fair brows to keep it on, his hands in his pockets,
the very picture of a careless saunterer, but a great wave of despair
was surging in his heart.
“My poor Neva!” he said to himself. “Where is she this wild day?
Does she begin to think I am never coming to rescue her?”
His wild glances, straying over the boats in the river, settled at that
instant upon a graceful yacht just coming to anchor. He could read
on her stern her name—The Arrow. He watched her idly for a long
time. He saw a boat lowered from her deck, and two sailors descend
into it. A gentleman in greatcoat and tall silk hat followed them, and
was rowed toward the shore.
The young earl started, his blue eyes flaming. Something in the
attitude and carriage of the gentleman excited his keenest scrutiny.
As the boat came nearer, and the faces of its occupants were
revealed more plainly, a strange cry sprang to Lord Towyn’s lips. He
had recognized in the tall stranger gentleman his own and Neva’s
enemy—Craven Black!
 CHAPTER XXI.
TROUBLE AHEAD FOR THE PLOTTERS.

It was a strange fatality that had brought Craven Black to Inverness

at the very time when his enemies were there, engaged in a search
for him.
The cold weather had set in early in his Highland retreat, and already
a foretaste of winter reigned in the Wilderness. Octavia’s cold had
settled upon her lungs, and she coughed dismally, and in a way that
made Artress tremble lest the wealth upon which they all counted
and Octavia’s life should die out together. Medicine was needed
imperatively, and so were all manner of winter supplies: and it was
for these things Craven Black had ventured from his stronghold like
some beast of prey, and when he should have obtained them he
intended to return to the Wilderness, and there wait patiently until
Neva should yield to his infamous demands that she should wed his
son.
He sprang nimbly from the boat on reaching the wharf, and taking
the cigar from his mouth, said, in a voice loud enough to convey his
words to the ears of Lord Towyn, whom he had not even seen:
“I am going up among the shops, boys, to make my purchases. I
shall go nowhere except to the shops, for I don’t want to incur any
risks. Be guarded, and say nothing to any one about your business
or employer. I sha’n’t go aboard under two or three hours, and you
can spend the time in some gin-shop. Here’s a crown to divide
between you, but don’t get drunk. Remember, we must get all our
stores aboard early and be off at daybreak in the morning. Be back
at the boat here in a couple of hours.”
The men assented and took the money given them, hurrying away.
Craven Black walked in an opposite direction, and was soon lost to
view.
Lord Towyn turned his gaze on the yacht. He saw that a man had
been left in charge of the yacht, for the captain stood on the deck
smoking a pipe.
A waterman was rowing along the river near the wharves, and Lord
Towyn signalled him. He came up swiftly to the edge of the wharf
upon which the young earl stood, and doffed his tarpaulin.
“Row me out to the yacht yonder,” said Lord Towyn, springing into
the boat. “I’ll board her on the further side.”
He sat down and the waterman pulled lustily out into the stream. In a
very brief space of time the boat had crept into the shadow of the
yacht on its further side, and the young earl climbed lightly to the
deck.
The captain approached him, scowling.
“Hallo, you there; what do you want?” he demanded gruffly. “This
here’s a private yacht, sir.”
“So I supposed,” said his lordship coolly. “It belongs to Mr. Craven
Black, doesn’t it?”
The pipe fell from the captain’s hands. He changed color.
“I don’t know as it makes no difference who it belongs to,” he said
blustering. “I work for pay, and it makes no difference to nobody who
I work for. Get out of this, young man!”
“Come, come,” said Lord Towyn sternly, his blue eyes blazing with a
light that terrified the captain. “None of your bluff, fellow. It won’t work
with me. I know with whom I am dealing. You are the servant of
Craven Black, and what have you to urge against my having you
indicted with him for the abduction of Miss Wynde?”
The captain fairly gasped for breath. He looked helplessly toward the
shore.
“I haven’t done nothing,” he muttered presently, quailing under the
steady, fixed and stern gaze of the earl’s blue eyes. “I’ve hired as
cap’n, and I am cap’n; but that’s all. On my soul, I an’t no party to no
abduction.”
“Your past character will be examined to throw light upon your
present motives and doings,” said the earl, with a telling random
shot.
The captain winced and quavered.
“Now see here, my man,” said Lord Towyn, following up his
advantage. “If you have done anything wrong in the past, it will be
like a leaden weight to drag you down when the officers of the law
arrest you for assisting in the abduction of Miss Wynde. It makes no
difference to you whom I am. Your game is up. The officers of the
law have accompanied us up from London. There are four of us in all
upon the trail of Craven Black. Refuse to do as I tell you, and I’ll call
yonder policeman on the wharf and give you up on the charge of
abducting a lady of rank and fortune.”
“What do you want of me?” the captain asked falteringly. “I haven’t
had my pay from Mr. Black, and he’ll kill me if I betray him.”
“He need not know that you have betrayed him,” said our hero. “You
need not leave this vessel. All that you have to do to secure your
safety is this: Tell me exactly the whereabouts of Miss Wynde, and
her friends will follow the yacht at a safe distance, quite out of sight
from your deck, and you can remain on board and collect your pay of
Mr. Black. He need not suspect that you have betrayed him.”
“And what will I gain by betraying him?”
“Your freedom from arrest. You see the policeman still stands on the
wharf? So sure as you refuse to speak, and speak quickly, I’ll call
him!”
Lord Towyn meant what he said, and the captain comprehended the
fact. He saw that Craven Black had been caught in a trap, and with
the usual instinct of villainy, he resolved to save himself from the
general ruin.
“You swear that I shall be protected if I tell the truth?” he asked.
“I do.”
“I suppose it’s penal servitude for abducting a young lady,” said the
captain. “Jack said so last night. Blamed if I haven’t been sick of the
job, anyhow. I don’t mind a mutiny at sea, when there’s cause, but I
can’t relish making war on a helpless girl, and I haven’t from the
first.”
“Come to the point. Where is Miss Wynde?”
“At a place up in the mountains owned by Mr. Black, and called the
Wilderness.”
“Miss Wynde is there, and Mrs. Craven Black?”
“Yes; and the servants, and Mrs. Artress.”
“How far from here is the Wilderness?”
“About twenty miles, if you could go direct, which you can’t. It’s forty
miles by water. You know Cromarty Frith, up in Ross—and—
Cromarty?”
Lord Towyn assented.
“You go about half-way up the Frith and turn into a river that leads up
among the mountains. The stream grows narrower as you ascend,
but the water remains deep, and at last your boat fetches into a
small loch lying at the foot of the mountains and surrounded by
them. A steep mountain rises right before you. Half-way up its side
lies a wide ledge, and on that the house is built. It’s a wild-looking
spot, sir, and a dreary one. It’s mighty cold up there, and I haven’t
relished the prospect of spending the winter there, not by no means.”
The earl asked several questions, to make himself conversant with
the route, and also to make sure that the man was not lying to him.
Assured on this point, he said:
“Craven Black intends to return home in the morning. He must not
suspect that you have betrayed him. Keep the secret, and we will do
so.”
The captain’s look of fear showed conclusively that he would be
careful not to allow his employer to suspect his treachery.
After a few further words, Lord Towyn re-entered his boat and
returned to the shore.
“Anything more, sir?” asked the waterman, as the boat ground
against the wharf.
“Nothing more—unless,” added the earl, with a sudden thought, “you
could direct me to a small vessel, a fast sailer, that I could hire for a
day or two. I should want a couple of men to take charge of her.”
“I don’t know of any such boat,” said the waterman, scratching his
head. “And yet The Lucky might suit you, sir, though it isn’t a
gentleman’s boat. She’s built for a fishing vessel, is brand-new, and
had a trial sail the day afore the gale, when she went like a bird.”
“I’d like to see her. Take me out to her.”
The Lucky lay out in the stream, half a mile further inland. Lord
Towyn rowed out to her, and found her joint-owners, two brothers, on
board. He went over the vessel, and found it new and clean, and in
fine order. The owners were willing to let the little craft with their
services, and the young earl hired it for a week, paying in advance
twice the sum the thrifty Scotsmen demanded for it.
“She must be provisioned immediately,” said Lord Towyn. “Her
destination is a secret, which I will tell you in the morning. I have
three friends who will make the excursion with me. We shall want
blankets, and all kinds of cooked meats and stores. We must leave
Inverness at day-break. Come ashore with me, one of you, and I will
select the stores we are likely to need.”
One of the brothers accompanied the earl ashore, and conducted
him to various shops, Lord Towyn keeping a keen look-out for
Craven Black, in order to avoid him.
Blankets and mattresses and bed linen were sent down to The
Lucky; various kinds of cooked meats, including rounds of roast
beef, roasted chickens, meat puddings, ham and veal pies, smoked
salmon and boiled ham, were packed in hampers and sent aboard;
and Lord Towyn added baskets of fruits, both dried and fresh, and
jams and confitures of every sort in abundance, besides boxes of
biscuits of every description.
“It looks like a v’y’ge to Ingy,” said Macdonald, the one of the two
brothers who had accompanied Lord Towyn ashore, contemplating
the array of stores with kindling eyes. “We can provision a ship’s
crew to Australy.”
“Whatever is left, you will be welcome to,” said the earl, smiling.
The young lord saw his new purchases deposited on board The
Lucky, and himself attended to the arrangement of the little cabin,
and then paying his waterman liberally, he returned to his hotel.
The day had passed swiftly, and he found that it was nearly five
o’clock of the short afternoon, and the street lamps were lighted,
when he entered his hotel and went up stairs, two steps at a time, to
his sitting-room.
 CHAPTER XXII.
THE FINAL MOVE IMPENDING.

The young Lord Towyn came into his sitting-room and the presence
of his friends like a sunny south breeze, all hopefulness and
gladness. He found Sir Harold walking the floor, his head bowed
upon his breast, his face ghastly pale, his eyes haggard, his mind
bordering upon distraction. The father’s anxieties concerning his
missing daughter was almost more than the overstrung brain and
tortured heart could bear.
Mr. Atkins stood at the window, sullenly despairing.
Ryan, completely at his wit’s end, sat before the blazing fire in a
crouching attitude, and with crestfallen visage.
Sir Harold, Atkins and Ryan turned, as by one impulse, toward the
young earl, as he bounded lightly into the room. Their eyes
expressed their wonder at the change a few hours had wrought in
him.
“You look like incarnate sunshine, Arthur,” said the baronet sadly.
“Ah, the elastic spirit of youth!”
The young lord laughed joyously, his fair face aglow with the
gladness that filled his being. He took off his greatcoat and hat, great
drops of sleet or rain falling from it, and shook himself, as he said:
“There’s a mist beginning to fall. We shall have a wet day to-
morrow.”
Sir Harold looked more keenly at the young earl, and a sudden
excitement possessed him.
“What has happened, Arthur?” he demanded. “You have news of
Neva?”
The glad smiles rippled like waves of sunlight over the young lord’s
mouth, and a joyous light danced in his blue eyes.
“Yes,” he said, “I have news of Neva.”
In an instant, Sir Harold, Atkins and Ryan crowded around our hero,
all eagerness and excitement.
“Is she here, at Inverness?” cried Atkins. “Have you seen her?”
“No; she is hidden among the Highlands, up in Ross and Cromarty.
She is safe and well.”
“But how do you know it?” demanded Ryan, full of professional zeal
and jealousy at once.
Lord Towyn took in his own the right hand of Sir Harold and pressed
it, as he answered gravely:
“I have seen Craven Black!”
A series of exclamations burst from the earl’s hearers. Sir Harold
Wynde trembled like a leaf, and turned his back upon the keen-eyed
detective.
“Black is here?” ejaculated Ryan, all thoughts of “the mysterious Mr.
Hunlow” lost in his sudden amazement. “Then our trip to Scotland
was not a wild-goose chase, after all? Tell me where he is, my lord;”
and Ryan took a step toward the door.
“Leave the management of the affair to me for the present, Ryan,”
said Lord Towyn. “You have placed us on the right track, and you
shall have all the credit for doing so, but we want no arrests, no
noise, no scandal. The matter must be hushed up for family reasons.
No whisper must go forth to the world of the wickedness of Lady
Wynde. The failure of the conspirators must be their punishment.”
“Yes, yes,” said Sir Harold. “The earl is right. There must be no
scandal.”
The detective shot a keen glance at the baronet, whose back was
still toward him.
“I have discovered,” continued Lord Towyn, “that Craven Black has
an old house up in the Highlands, and that he has Neva there in
close imprisonment. Neva has all the courage for which we gave her
credit, and holds out stoutly against her enemies. Black apprehends
a long stay in the Highlands while reducing her to subjection, and
has come to Inverness to-day for winter stores. His yacht lies in the
river, and he will be off at daybreak.”
“What are we to do?” exclaimed Atkins. “He will escape us. It will be
necessary to arrest him.”
“I have made all the preparations for pursuit,” replied our hero. “I
have a sloop, provisioned and ready, in which we will follow the
yacht. We will dine here, and at a late hour this evening we will go
aboard our vessel. We must be off in the morning, soon after the
yacht.”
Sir Harold plied the earl with questions, and the latter told the tale of
his day’s adventures at full length.
At six o’clock, dinner was served in their sitting-room. After dinner
they talked for hours, and at ten o’clock that evening they quitted
their hotel and went down to the wharf.
A fine mist was falling. The river was shrouded in darkness, but
watery gleams of light came from the yacht and the other vessels
lying in the river. The wind was still strong and free, blowing in fitful
gusts.
The party had hardly come to a halt upon the wharf, when a small
boat manned by two rowers shot out of the gloom, and ran along-
side the landing-place.
“Is it you, Macdonald?” asked the young earl, bending forward and
straining his eyes through the darkness.
“Aye, aye, sir. Is it you, my lord!” said Macdonald, in a loud whisper,
rising in the boat.
“You are on time,” said Lord Towyn. “The clock is striking the hour at
this moment.”
The earl assisted Sir Harold into the boat, and sprang in after him.
Atkins and Ryan followed, and the boat shot out again into the
stream. A vigorous row of several minutes on the part of the brothers
Macdonald, brought the party alongside The Lucky, and they
climbed to her deck. The boat was drawn up and secured to the
davits.
The young earl led the way out of the wet and gloom down into the
cabin. There was no fire here, but a lantern hung from the ceiling,
shedding a bright light upon the table and the bench around it. The
air was chilly, and the small cabin, despite its snugness, had the
comfortless appearance usual to a ship’s cabin.
“It’s cold here,” said the elder Macdonald, the captain of the sloop.
“I’ll just mix you a punch. There’s nothing like a good punch to keep
the rheumatiz out of one’s bones.”
He brought from the pantry a spirit lamp, which he lighted. Over this,
upon a light frame-work, he placed a tea-kettle. Then he brought out
an immense bowl and ladle, a netted bag full of lemons, a tin can of
sugar, various spices, and a jug of Scotch whisky. When the water in
the tea-kettle boiled, he rolled up his sleeves and set to work,
concocting a punch which proved a success. The four passengers,
as well as the two Macdonalds, did full justice to the warming
beverage.
About eleven o’clock Sir Harold Wynde was persuaded to retire to a
state-room and berth, but he did not sleep. A host of anxieties
pressed upon him, and he tossed for hours on his hard bed, the prey
of a torturing anguish.
Atkins and Ryan went to bed, and to sleep.
Lord Towyn went out upon the deck, and walked to and fro with
Macdonald, whom he found to be a shrewd, kindly Scotsman, and
he told him enough concerning the misdeeds of Craven Black, and
the abduction of Miss Wynde, to interest him personally in the
restoration of Neva to her friends.
“We must keep an eye upon the yacht all night,” said the earl. “She
may creep out in the darkness, and if she attempts that, we will
creep after. I know the route she will take, but I prefer to arrive at the
inland loch not an hour behind her.”
“I will keep watch,” said Macdonald. “You ought to get a little sleep,
my lord.”
“I cannot sleep,” answered Lord Towyn, more to himself than to his
new ally. “My veins seem full of quicksilver to-night, rather than full of
blood. So near to her—so near!”
He paced the deck alone in the mist and the darkness. Around him
the watery gleams of light flickered upon the river, and from the
wharves other watery gleams strayed, looking like reflections of the
first. The bells rang the hours with muffled clangor. A strange
ghostliness fell upon the dark river and the sleeping city.
All through the night Lord Towyn continued his weary tramping to
and fro upon the deck. One of the Macdonalds shared his vigil; the
other slept.
The Arrow lay quietly at her moorings throughout the night, but at
daybreak signs of activity were seen upon her deck. The morning
had not fairly dawned when the yacht slipped out of the stream,
heading toward Moray Frith.
The breeze was favorable to her progress, without being astern, and
she bowled away at a fine rate of speed. The young earl, looking
through a sea glass, could distinguish four figures upon the deck of
The Arrow, and one of these he recognized as that of Craven Black.
“Shall we shake out the sails, my lord?” asked the elder Macdonald.
“The yacht goes well, but I fancy we might show her the heels of The
Lucky in a fair race. We can keep her in sight the whole distance.”
“That won’t do, Macdonald. Black must not suspect he’s followed.
This mist is like a vail, and will conceal us if we keep at a reasonable
distance behind. Let him get half way to Fort George, and then we
will start.”
In good time, when The Arrow had made about the distance
indicated as desirable, The Lucky slipped from her moorings, and
shaking out her sails as a bird shakes its wings, she flew onward
over the waters in pursuit.
The sloop was half-way to the fort, and the yacht was out of sight,
when the baronet, fully dressed, and showing by his haggard face
that he had not slept during the night, came out upon the deck.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com