SQME VARIABLES BF IMPORTANCE .
[N DETERMlNlNG REPORTS
3F DIREC?EQNAL APPARENE MBVEMERT

Thesis for the Degree of Ph.D.
mam EM??? iféi‘c’EREi‘fY

Martin Krampen
1962

It-i £125.18

This is to certify that the

thesis entitled

Some Variables of Importance in Determining Reports
of Directional Apparent Movement

presented by

Martin Krampen

has been accepted towards fulfillment
of the requirements for

Ph D degree in Communication Arts

 

)V 0 O <§flk. I.

Date 5-18-_62

 

0-169

 

 

 

APR032015
100815

 

ABSTRACT

SOME VARIABLES OF IMPORTANCE
IN DETERMINING REPORTS

OF DIRECTIONAL APPARENT MOVEMENT

by Martin Krampen

In a symmetric configuration of apparent movement
stimuli, SS tend to see bilateral apparent movement, when
the light in the center is flashed first and rapidly followed
by the two equidistant lights at both sides. This phenomenon
is called "tendency to total assimilation!‘ The hypothesis
of this study was the following: Exposing S8 to active
practice with one direction in the symmetric configuration
(flashing first the center light and then one of the side
lights) will increase on the posttest the number of reports
expressing a directional preference (if such preference
is sollicited). The increase of preferential reports is
expected for the direction which was practiced.

After §s had been tested whether they saw apparent
movement at all, they were exposed in a pretest to nine
identical trials of the complete symmetric configuration.
§s were sollicited to detect differences in the symmetrical
pattern. The pretest was used to set a level of initial
performance and to sort out §S who were directionally

biased to begin with.

Martin Krampen

§s were then randomly assigned to one control and
two experimental groups. One experimental group practiced
apparent movement to the right, the other to the left.

Both groups reinforced their directional experience by
making lever movements in the direction of practice and
pronouncing directional phrases. The control group re—
ceived a time estimation task instead of directional
practice.

On the posttest gs were again exposed to the nine
identical trials of the pretest. Again detection of dif—
ferences in the symmetrical pattern was sollicited.

A posttest check, an interview, and a questionnaire
were used as controls,(for example, one would like to control
whether gs continued seeing apparent movement throughout
the experiment and whether they were aware of E's hypothesis)-

The data consisted of the number of preferential
reports in the left or right direction. If there was an
excess of more than one third of nine reports with specific
directional connotation on the pretest, § was defined as
biased.

A change score between pretest and posttest was
computed by subtracting the number of preferential reports
on the pretest from those on the posttest. The significance
of the (mean) change between experimental groups and controls

was tested by‘t test and chi square.

Martin Krampen

It was found that §s with practice increased sig—
nificantly the overall number of reports with directional
connotations when they were not biased, whereas no such
increase could be observed in biased.§s and controls. The
variability seemed to increase in all groups.

Problems of interpretation of the results arise
from the fact that unilateral practice did not increase the
number of reports with connotations in the direction which
was practiced.

While completely unbiased.§s reacted in the direction
opposite to their practiCe, slightly and strongly biased
§s alike did not significantly exhibit such counter reaction.

When §s were grouped in terms of their initial bias,
it was found that practice did increase connotative responses
in the direction of practice only when it worked against
the direction of initial bias. When initial bias was
reinforced by practice §s acted against the direction of
practice.

All groups exhibited a significant bias reversal
from pretest to posttest. Practice thus interacted with

bias reversal to produce the complexity of the above results.

SOME VARIABLES OF IMPORTANCE
IN DETERMINING REPORTS

OF DIRECTIONAL APPARENT MOVEMENT

by

Martin Krampen

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Communication Arts

1960

ACKNOWLEDGMENTS

The planning and completion of this study would have
been impossible without the patient guidance and help of
my entire committee at Michigan State University:
Dr. S. Howard Bartley, Department of Psychology
Dr. Paul J. Deutschmann, Head, Communications Research
Center
Dr. Malcolm S. MacLean, Jr., Acting Head, Communications
Research Center
Dr. Hans H. Toch, Department of Psychology
Dr. William T. Stellwagen, Departments of Communication
Arts and Psychology.
To all of them I wish to express sincerely my gratitude.
I am also indebted to friends at Michigan State University
and colleagues at the Department of Psychology, Carnegie
Institute of Technology, for the encouragement gathered
from many discussions about the methodology and theory of
the present study. Much of the preliminary work and the
experiment itself were carried out under N.S.F. Grant G-86lO,
headed by Dr. Toch. The scope of this grant was to in—
vestigate the relationship between "meaning" and apparent
movement. Miss Anne Beard, editor in the Communications
Research Center, Michigan State University, was kind enough

to read and improve the manuscript.

ii

Table of contents

ACKIJOWIEDGIEIP‘ITS O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 0 O O O i i
Lj-St Of tables and figures 0 O O O. O O O O. 0 O O O O 0 O I O O O O O O O O O V
Appendices O O O O O O. 00 .0. O O O O O O O. O O O O O O. O O 0.... 0.. O O O O. O Vii

1. Introduction

1.1 Problem .................................. 1
1.2 Physiological hypotheses under criticism . 2
1.3 "Looking into the past" of the organism .. 6
1.4 A re-definition of the role of past experi—

ence in apparent movement ................ 11
1.5 How can "meaning" be built up? ........... 14
1.6 Some final caitions ...................... 19
1.7 The hypothesis ........................... 21

2. Method

2.1 Apparatus ................................ 23
2.2 Stir-{11,111 o00.000.000.00.00000000000000.0000 27
2.3 Subjects ................................. 28
2.4 General procedure ........................ 29
2.5 Analysis of data ......................... 37

3. Results

3.1 Biased and unbiased.§s ................... 45
3.2 Pretest and posttest change in directional
connotation ............................. 45
3.3 Pretest to posttest change with specific
directional connotation .................. 48
3.4 Pretest to posttest decrease of bias ..... 57
3.5 Homogeneity of the groups ................ 59
3.6 §s seeing through the experimental design 60
3.7 Attitude ................................. 61
3.8 Posttest check ........................... 65
3.9 ‘Ss reporting apparent movement ........... 65
3.10 Kinds of apparent movement experience .... 66
3.11 Kinds of "connotations" reported ......... 67

4. Discussion

4.1 Interaction of practice and bias ......... 69

iii

Practice ................................
The test ................................
The role of awareness ....................
Economy of experimental design ...........
Definition of connotation ................

Pete:
O‘KJ‘l-F'WN

summary OOOOOOOOOOOOO0.0......OOOOOOOOOOOOOOOOOOOO

Bibliography 0..0.00.00.00.00.0000COOOOOOOOO..0...
Appendices:

Appendixl OOOOOOOOOOOOOOOOOOOOOOOO00.00.000.001...
Appendix II OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

iv

72
H

W
W

82
85

90
W

IO.

11.

12.

List f tables and.figures

 

 

Title

 

Number of connotations to be reported
on pretest by relatively unbiased §S.

Pretest to posttest change in number
of reports with directional con—
notation for unbiased and biased.§s.

Summary of chi square and t tests
for differences between groups.

Central tendencies and variability
of the number of directional con—
notations reported by unbiased and
biased.§s.

Increase of number of reports with
specific directional connotation.

Interaction of practice and bias.

Posttest increase of unbiased vs.
biased.§s.

Posttest increase of unbiased §S
reporting directional connotations.

Posttest increase of biased SS
reporting directional connotations.

Increase in connotations to the
right, §s with practice to the
right vs. §s with practice to the
left.

Posttest increase of unbiased.§s
reporting connotations to the left
after right practice.

Posttest increase of unbiased §s
reporting connotations to the right
after left practice.

47

49

5O

51
56

99

100

101

102

103

104

 

15.

16.

17.

18.

19.

20.

21.

Figure 2Q,

Interaction
in slightly

Interaction
in strongly

Interaction

Title Page Q9,

of training and bias
biased Ss. 105

of training and bias
biased.§s. 106

of training and bias

in all biased gs. 107

Proportion of completely unbiased
and slightly biased §S in the
relatively unbiased experimental

 

groups. 108

Proportion of completely unbiased

and slightly biased.§s in the

relatively unbiased experimental

and control groups. 109

Difference in attitude between

experimental and control groups. 110

Difference in attitude between

experimental groups. 111

Difference in attitude between.§s

practicing to the left and §S without

practice. 112

Posttest check. 113
Title Page 2;-

Dark alley, viewing box, and lever

apparatus. 26

vi

Appendix I.

Appendix II.

Appendices

Specific Procedure and Instructions.

Chi Square Tables.

vii

9O
99

1. Introduction

1.1: Problem. In a symmetric configuration of apparent
movement stimuli, when one light in the center of a screen
is flashed and is followed rapidly by the flashing of two
equidistant lights at both sides, §s frequently report
bilateral apparent movement. The hypothesis of this ex—
ploratory study was the following: Exposing §S to active
practice with one direction of the symmetric configuration
(flashing at first the center light and then one of the lights
at the side) will increase the number of reports of direc—
tional preferences in the bilateral apparent movement
configuration. gs were given a pretest to find out whether
they were initially biased for one or the other half of the
symnetric pattern. An apparatus was devised to allow §s
active practice while viewing one or the other half of the
configuration. Then, the symmetric configuration was
repeated in the posttest, and an attempt was made to assess
whether practice had produced any change in perceptual
reports.

The purpose of this introduction is to sketch the
historical background of apparent movement research as it
relates to the development of perceptual theory. Perceptual
theory has developed from the earlier controversy between

nativists and structuralists on one side and empiricists

on the other side to a more streamlined and synthetic

view (Allport, 1955: pp. 86—89, 299—303). This develop—
ment has brought about the theoretical framework for apparent
movement research from which the basic hypothesis of this
study was deduced.

The theoretical framework suggests that "meaningless"
physiological receptor events get their "interpretation"
through interaction with stored traces of past experiences.
The present study is an attempt to test the plausibility

of such a statement in the area of apparent movement.

1.2: Physiolggical hypotheses under criticism. Perception
used to be regarded as a "stimulus—bound" functioning of
innate mechanisms. The argument for this position went as
follows: If a percept was not immediately given and veridi—
cal, the contact of the organism with the environment
would have to be thought of as maladaptive; this concept is
inconsistent with the functional point of view. Early
perceptual investigations thus usually centered around
certain aspects of perception with the intent of specifying
their mechanism. Perception of space, for example, was
examined in order to gain insight into the underlying
physiology.

Another aspect of perception, the perception of
movement, was linked to the beginnings of the Gestalt
school. Wertheimer's classical paper (1912) focused on

perceived strobosOopic movement and proposed a physiological

"short circuit theory” as explanation.

Other neurophysiological theories such as K8h1er's
StromfAdensaulentheorie were advanced later.

"Premature" neurophysiological hypothesizing in
explaining apparent movement (especially Wertheimer's
short circuit theory) has since been under criticism from
several quarters. One kind of criticism comes from neuro—
physiologically oriented experimenters.

Bartley (1941) has summarized the evidence concerning
the physiological mechansims underlying apparent movement.
Differential spread of activity in the retina furnishes a
physiological counterpart to reports of different kinds of
apparent movement. But, while they could not exist without
the retinal distribution, not all forms of apparent movement
depend to an equal extent on retinal factors.

The retinal conditions set the stage for
movement itself, while in complex stimulus patterns
the choice of which of the two directions it will
be is effected elsewhere. (Bartley, 1941, p. 162f).

Where then is the locus of apparent movement? According
to Wertheimer's theory, one would have to postulate a
unified field in either of the cerebral hemispheres.

K. U. Smith (1940) has demonstrated that head nystagmus
occurs as a response to both "real" and stroboscopic move—
ment stimulation in guinea pigs before and after removal of
the occipital lobes, hemidecortication, and complete bilat-

eral decortication. In another extirpation experiment

K. U. Smith and Kappauf (1940) found persistence of optic
nystagmus in cats after bilateral removal of the striate
cortex, again under "real" and stroboscopic movement con—
ditions. These authors conclude that the neural basis of
apparent movement seems to be located in subcortical optic
centers.

Shapiro (1954, p. 54) has criticized this view by
calling attention to the possibility that in man, as opposed
to lower vertebrates, the cortex may have taken over all
optic functions (with the exception of the pupillary reflex
to light).

An investigation of Teuber and Bender (1949, p. 41)
using eleven war patients with injuries of the occipital
lobes as SS, seems to confirm Shapiro's criticism. The
results showed for all §s a conspicuous reduction in the
ability to perceive apparent movement.

Since "satiation" (in the K8h1er—Wallach sense)
would reduce conductivity in the brain field (as would be
the case in injured brain tissue) apparent movement should
be found reduced after continuous stimulation too. Evidence
for such a satiation effect was found by Deatherage and
Bitterman (1952) and Shapiro (1954), whereas Brenner (1953,
1956) obtained results contrary to a satiation hypothesis.

Gengerelli (1948) found apparent movement reports
with homonymous and heteronymous stimulation of the cerebral
hemispheres. His technique was to direct the first of

two flashes into one retina, the second into the other

retina. A similar technique was used by K. R. Smith (1948).

He fastened two punctiform lights to the head at the temporal
margin of the eye. Either light was visible only to its
corresponding eye at the nasal end of the horizontal median.
Assuming that there was no important neural interaction in
the region of the chiasma, he concluded that he was not
stimulating a unified neural field which would be implied

by the short circuit theory. Nevertheless, seven out of

ten of his trained observers reported apparent movement,
mostly of the "phi" variety. A review of neurophysiologically
oriented experimentation shows that at the present it is

not possible to characterize precisely the neural process
underlying apparent movement.

Another criticism of early neurophysiological theories
came from investigations stressing the influence of "set."
Various apparent movement experiences could be elicited
under the same extra-organic conditions. Such findings
soon led to scepticism toward exclusively neural hypotheses,
especially in the United States. According to Neff (1936,

p. 29), Titchener delegated one of his students, F. L.
Dimmick, to investigate critically Wertheimer's theory
(Dimmick, 1920). Dimmick and Sanders (1929), De Silva
(1926), Higginson (1926), Kelly (1935), and in Europe,
Neihaus (1930), Pieron (1933), and others demonstrated

that different "sets" or "attitudes," usually defined
operationally as "different instructions," produced different
movement experiences, although all external conditions

were kept constant.

A further criticism was based on the argument that
responses acquired in the past may interact with perception
of apparent movement. How, and to what degree, this inter-
action takes place is not yet exactly known. That this
question needs careful investigation was pointed out by
Neff (1936). Neff discusses the major studies in apparent
movement, beginning with Exner's article of 1875 and ending
with studies from the early thirties. Then he reviews
the variables investigated and the theories advanced. In
his conclusions, he recommends the rejection of "mental
entities" and premature neural hypotheses as explanations
of apparent movement. He sees two major problems as guide-
posts for further research:

The first of these problems is to give a
suitable account of the conditions (within and
without the organism) which must be satisfied at
the moment of perceiving. The second problem is
concerned with the development of the performance.
The fact that, in the adult human organism, percep—
tions of depth, form, color, movement, rate, change
and other aspects of objects and occurrence may be
called forth by a variety of conditions; and that,
moreover, the same extra—organic conditions may
lead now to one, now to another variety of perceiving,
leads us to suspect that we must look into the past
to complete our account -—we must point to the ob—
vious fact that earlier performance exerts some
influence upon later performance.... (Neff, 1936,
p. 41)

1.3: "Looking into the past" pf the organism. One way of
"looking into the past" of organisms would be to adopt a
developmental approach. Is apparent movement different in

children and adults? Meili and Tobler (1931) studied

apparent movement in children. They found that children

have a lower apparent movement threshold than adults.
Brenner (1957) repeated the experiment of Meili and Tobler.
She studied 111 SS from two and one half to 19 years old.
The results of Meili and Tobler were, in essence, confirmed:
Changes in threshold of apparent simultaneity and apparent
movement were clearly a function of age. Both illusions
were seen under objectively more difficult conditions
(slower flash frequencies) by younger children. Brenner
also found that the capacity of perceiving apparent movement
and simultaneity decreased with age until a level was
reached at 14 years. She attributes this detriment to an
increase in discrimination capacity with age as against
the explanation of Meili and Tobler that children form
"Gestalts" more readily than adults. Another developmental
study was carried out by Gantenbein (1952). She too found
that the younger the children were, the more apparent movement
they saw. For instance, younger children had apparent
movement experiences of subjectively longer duration.

Werner and Thuma (1942) studied apparent movement
in brain-injured children. Brain—injured children differed
from their controls by failing to experience apparent
movement although they were able to perceive some movement
when the stimulus figure was "meaningful" or when they

looked at "real" movement.

The extent to which earlier performance influences
movement perception even in children is difficult to assess.

A structuralist could maintain that at the age in which

apparent movement is studied in children the mechanism of
perception has already matured to a certain degree. An
empiricist, on the other hand, could postulate that movement
perception habits are formed at an early age. It is obvious
that, for our question of present performance as influenced
by past performance, the developmental approach presents
many difficulties.

Another way of "looking into the past" is through
the use of "meaningful" stimulus material. The assumption
in this approach is that meaning connotations of an object
are the cumulative result of past dealings with that object.
As Hastorf (1950, p. 213) puts it: "Attributing meaning is
essentially the application of past experience to stimuli."
In principle, then, it would be possible to test the in-
fluence of past performance upon present performance by
comparing responses to a meaningful and a meaningless
stimulus. We would expect apparent movement to be "dif-
ferent" with meaningful material when compared to a meaning—
less stimulus. The latter should trigger the perceptual
mechanism in its purest form. It is difficult, however,
to conceive what a strictly meaningless stimulus might be.

Meaningful stimulus figures have been used in apparent
movement study from its very beginning in the 19th century.
However, De Silva (1926, pp. 473, 480, 497) must be credited
with having first demonstrated that meaningful figures give
a clearer perception of movement than abstract patterns.

Steinig (1928, p. 324—326) used pictures of familiar objects

in an investigation on the perception of the passage of
apparent movement between two end points. Blug (1932, pp.
292-297; 300—305) found that meaningful material not only
increases the degree of apparent movement, but also con—
fers directional connotation to the movement. Meaningful
stimuli were also used in the developmental studies of
Meili & Tobler and Werner & Thuma. (This detail is not
mentioned in the discussion of these studies and may indeed
add a new dimension to their interpretation.) Next, it was
Pieron (1933, pp. 246, 248) who used meaningful figures in
a critical investigation of the structuralist position. He
stressed general attitude and "experience" as important
determinants of apparent movement. Jones and Bruner (1954)
compared the quality of movement in nonsense figures and
meaningful stimuli. After the first trial, 15 of their 18
SS reported that the figure of a running man moved faster
than a comparison stimulus featuring a nonsense object.

On the fifth trial, however, eight §S saw the man and the
nonsense object moving at equal speed, and eight of the
remaining ten saw the nonsense object moving faster and
further. Jones and Bruner suggest that:

Expectancy operates to accentuate the per—
ceived velocity and distance of the man's movement
relative to that of the nonsense object on the first
trial when the actual spatial—temperal features are
still equivocal, and when the objects are not yet
clearly localized within a stable framework. When
the framework is established there appears to be a
tendency toward reversal. We do not know how to
account for this effect save to suggest that Kbhler—

Wallach (1944) satiation may be operative in apparent
movement. (Jones & Bruner, 1954, p. 159).

IO

Brenner (1956a; 1956b) used a geometrical angle, an abstract
design, the picture of a stick man and one of a pendulum
clock as stimuli in her experiment on meaning and apparent
movement. She labelled three types of apparent movement.
The first type she called after Dimmick and Sanders (1929)
"visible" apparent movement. Visible apparent movement is
experienced when under certain conditions of duration and
interval stimuli are flashed in a repeated cycle. It is

not affected by meaningful stimulus figures. The second
type of apparent movement was labelled "inferred" movement.
According to Brenner, inferred movement is a reconstruction
after the completion of the perceptual event. The object

is first seen in one position and then in another. §
"interpolates" the movement or "rationalizes" the change of
the object's position by concluding that "it must have
moved.” The "interpolated" variety of inferred movement

is produced when stimuli are flashed in a repeated cycle

(as in the case of "visible movement") but at time intervals
over 200 milliseconds. The "rationalized" variety of
inferred movement is reported when only a single presentation
of stimuli is used. While the effect of meaningful figures
on "interpOlated" movement is still unknown, Brenner finds
that meaning plays a major role in determining the upper
threshold of "rationalized" movement: The symbolic rep-
resentation of the stimulus in terms of a moving figure

enhances the facility to rationalize movement.

11

1.4: 'A e—definition.p£ the role pf past experience_;n

 

apparent mgygmggg. Toch and Ittelson (1956) wrote an answer
to the Jones and Bruner study. They took this occasion to
review carefully the literature concerning the role of

past experience in apparent movement. First they outlined
the history of experimentation with meaningful stimulus
material. Then they reviewed this type of experimentation
in a larger theoretical framework, that of the debate between
empiricism and structuralism. By quoting from Wertheimer
and Metzger, they showed that even the Gestaltists were
aware that "past experience" is a variable which merits
consideration (although its importance is minimized by the
argument that experience could never be gained without the
operation of Gestalt mechanisms). Toch and Ittelson then
continued:

In view of this position of Gestalt theories,
and since no empiricist would deny the physiological
substratum of perception, the fundamental question
has been reduced to one of relative primacy or‘
importance of learning processes and physiological
mechanisms instead of one of exclusive determination.
(Toch & Ittelson, 1956, p. 198)C

They even conceive of the possibility that the experimenta—
tion with meaningful material might not be decisive for the
isolation of the role of past experience. For an empiri—
cist such experimentation might simply put in synchronism
or antagonism two relative weights assigned to different
experiences. One would be the rather deep-seated and

heavily weighted assumption that rapid succession in stimu—

lation of different locations in the retina corresponds to

12

"movement." The other would be a less heavily weighted
assumption constituted by a particular meaning (e.g., direc—
tional connotation) connected to the stimulus figure.
Structuralists, on the other hand, could use experimentation
with meaningful stimuli as a means for testing to what
extent the operation of the physical mechanisms of perception
can be modified through "experience."

This moderate and refined empiricist position pro—
posed by Toch and Ittelson requires a particular experi—
mental design to be tested. In order to find out whether
meaning is an effective variable, it must not be brought
into antagonism with more powerful determinants. Rather,
where apparent movement could occur without contradiction
in several directions, one of these directions should be
accentuated by meaning connotation; the test would be
whether the movement in the accentuated direction is in
some way perceived as "outstanding."

Following this rationale, Toch and Ittelson adapted
a design used by Von Schiller (1933) for their purposes.

Von Schiller's question, from a structuralist's point of
view, was the following: Given a stimulus configuration

in which alternative apparent movements could occur, which
of the alternatives will be perceived by S2 The hypotheti—
cal answer was that normally movement would be expected to
occur in all of the possible alternative ways. Von Schiller
called this a "tendency to total assimilation." Part of

his experimentation was designed to explore how the manipu—

13

lation of structural variables (form, color, configuration)
would alter this tendency. The particular design which
Toch and Ittelson adapted was used by Von Schiller (1933,
p. 185) in an exploratory study. A stimulus figure a was
first flashed in the center of the other equidistant stimulus
figures b1 and b2. The question now became: Given the
normal tendency to total assimilation in apparent movement,
and keeping all structural variables constant, will meaning
connotation be effective in offsetting total assimilation?
For example: If the same figure of a jumping deer were
flashed in a b1 -ar b2 sequence, would §_perceive the
movement in the direction of the jump as "outstanding?"
Further questions arise immediately, e.g., what kinds of
outstandingness would be experienced? Toch and Ittelson
in their exploratory study used figures of a bomb and an
airplane as stimuli with movement connotation; the control
stimulus was the figure of a bottle, presumably without
movement connotation. The results were in the expected
direction. Movement connotation was effective in unbalancing
total assimilation. Unilateral movement was reported either
exclusively or with preferential expressions; with the
control stimulus, instead, the tendency to total assimilation
prevailed.

An experiment by Krampen and Toch (1960) followed
essentially the same line of reasoning set forth in the
exploratory study of Toch and Ittelson. The stimulus

figures used were arrow designs ranging from a "perfect"

l4

conventional arrow figure through several figures with
decreasing "arrowness," to a completely symmetrical shape,
similar to a barbell. "Arrowness" was decreased by mutila-
ting or substituting such features as the tip and the tail
of the arrow. In essence, the hypothesis of the effect

of meaning in Von Schiller's total assimilation situation
was confirmed. The effect was strongest for the "perfect"
arrow design, but no regularly decreasing effect could be
established for the figures with decreasing "arrowness."
Directional connotation in arrow designs seemed to be
somewhat an "all or none affair" depending on certain
features (e.g., tip and tail) which mediate the recognition

of the stimulus as an arrow.

1.5: Hpfl pap Tpeaning" pg ppglp pp? In summing up, it
seems that both structuralists and empiricists would agree
on the proposition that "acquired responses" or "past
experience" enter in some way into the functioning of physio—
logical perceptual mechanisms. The question of the "how"
and the degree of interaction between past experience and
apparent movement is still not answered. Experimentation
with meaningful material can only establish whether there

is or is not such an interaction in the presence or absence
of other structural variables. The present exploratory
study was devised to obtain more information on how this
interaction actually occurs. If "meaning" is an application

of past experience to stimuli, and this experience interacts

15

with perceptual mechanisms, then it should be possible to

bring this process under control in a laboratory situation.
The first step would be "building up" a meaning

(or connotation) for a previously relatively meaningless

stimulus pattern. Using the Von Schiller design (b1-§_

b2), one could for instance flash repeatedly a unidirectional

apparent movement (e.g., g-bg) by omitting one of the b

stimuli. The assumption would be that repeated presentations

of a unilateral g—b sequence would be one possible method

of developing meaning: S would perhaps experience such a

sequence with some directional connotation. The second

step then would be a test of whether meaning formation has

in fact occurred. It would consist in presenting the full

Von Schiller bl—g-bz pattern. Will S now perceive one

part of the pattern endowed with some directional connotation?
There are at present several theories about how

”meaning" can be built up. Some of these theories are

reviewed by Osgood (1952, 1958) and Osgood, Suci, and

Tannenbaum (1958). These authors refer to the naive

application of Pavlovian conditioning principles as the

”substitution theory” of meaning. Another theory (Ogden

& Richards, 1923) holds that "ideas" are the essence of

meaning. This theory is labelled by the above authors as

"mentalistic." The same criticism is made against an early

theory of Charles Morris (1938). In this theory Morris

states that signs achieve their meaning by arousing respon—

ses that "take account of" the objects they refer to.

 

On a la

16

In a later theory, Morris (1946) modified his view, now
the meaning of a sign consisted of a set or disposition
to respond to it with any of the responses previously
elicited by the object itself. Osgood himself proposes the
hypothesis that meaning is a mediation reaction, which
consists of some fractional part of the total behavior
elicited by the object, and produces self—stimulation.
The self—stimulation mediates responses which would not
occur without previous association of nonobject and object
patterns of stimulation.

The apparent movement study of Toch and Ittelson
(1956, p. 199) was based on a transactional approach to
meaning: Meaning is built up by looking at and dealing
with a stimulus pattern. Repeated exposure to that pattern
leads to the development of assumptions and sets. These
"experiences" are brought into the experimental situation
and are considered determinants of the response of record.

Since the aim of the present study was to produce
experimentally an influence of "past performance" on "present
performance" in perception of apparent movement, a com—
bination of various meaning-building techniques seemed
justified in order to make the "past performance" as im—
pressive and relevant as possible. Thus S should not only
look at a unidirectional apparent movement during practice.

He should make motor and verbal responses as well.

17

Such a methodological decision would find the approval
of transactionalists. For the pretest period of Kilpatrick's
experiment on "two processes in perceptual learning" (1954)
Ss viewed two distorted and one normally shaped room monocu—
larly and did not see the distortions. During practice,
one group was allowed to throw rubber balls and manipulate
a stick in the interior of one of the distorted rooms.

The other group viewed the same room passively, while E
manipulated the stick and threw the ball for S, On the
posttest both groups reported alterations in the appearance
of the distorted room in which they had practiced (e.g.,

in slope of the floor, etc.). These alterations in appear-
ance carried over to the other (not practiced) distorted
room and to the normally shaped room. Perceptual alteration
was slightly greater in the active than in the passive group.
In active §s alteration occurred in greater amount and

there was somewhat more carry-over to the second room
(distorted or normal). Although the difference between the
two groups was only slightly in favor of active practice,

the experimenter saw no reason to reformulate the role of
action in perception as it had been proposed by transactional
functionalists. He writes:

The author and others with a similar point of
view have stated that actions and their consequences
are instrumental in the "construction" of what we see.
What exactly is meant by action has not been well
specified, but the implication has always been,

probably intentionally, that it means gross overt
bodily activity, the doing of something....

18

Apparently, perceptual learning requires that
a relevant set of sequential events be set in motion
in such a way that successive impingements on the
learning organism will be attended to and related
over time to one another, to expectations, and to
purposes. Only in rare circumstances, such as in
an experiment, will these conditions even be approxi-
mated unless the learner is the operator. (Kil—
patrick, 1961, pp. 176, 185).
Other support for the design of the present study
comes from the theory of "active participation." This
theory grew out of a series of studies on affective pro—
cesses. Haggard and.Rose (1944) and Haggard and Babin
(1948) used active participation in the conditioning of
autokinetic movement. It consisted of drawing arrows by
pressing the pencil hard through several layers of carbon pa—
per. For example, if S saw movement occurring to the
right, he drew an arrow pointing to the right; if movement
was perceived to the left, the arrow had to be drawn pointing
to the left. Dependent variables in the first study were
frequency of movement, direction of movement, magnitude
(estimated in inches) of the movement, degree of confidence
for seeing movement, and stability of the CR. "Active"
SS saw the light move more frequently and farther. They
were more confident of their perception, and were conditioned
more stably. In the second study, one group of S; drew
arrows pointing in the direction of the autokinetic move—
ment. Another group drew arrows pointing in the direction
opposite to the autokinetic movement. Further variables

controlled in this study were reinforcement by reward and

eye dominance. Effectiveness was expressed in terms of a

19

complex index of conditioning. Most effective was active
participation (drawing arrows) in the direction of the
autokinetic movement. Second most effective was active
participation in the direction opposite to the autokinetic
movement. No active participation was least effective.
Both reward and eye dominance were less effective than
active participation.

It is interesting to note that even active partici—
pation in the direction opposite to perceived autokinetic
movement was effective in making §s perceive autokinetic
movement more frequently, of greater distance, and with
more confidence. This indicates that active participation
as such (even without directional synchronism) may be
instrumental.

Further support comes from Von Schiller's (1933,

p. 196) experiment. He found that a change in the direction
of movement ("Kippenlassen") was accomplished most success-
fully, not when the new perception was visually imagined,
but when an effort was made to experience it in terms of
motor adjustment. SS reported control over directional

reversals by means of motor imagery.

1.6: Some fina_ pautions. The meaning constructed in the
present study had to be of a directional nature. ‘Ss who
supposedly follow at first the tendency to total assimila—
tion are expected to change their perceptual performance

into a directional one because they have been exposed to

20

directional practice. To be able to say that this per—
ceptual change really has occurred, we must make the assump-
tion that S starts "from scratch," that is, that S has at
first the tendency to total assimilation. This assumption
can only be made with caution. It is conceivable that S
comes to the experiment with acquired directional biases
of a heavy weight. Von Schiller himself talks about a
"tendency" to total assimilation, which means that we have
to expect deviations from this tendency. That there are
directional preferences in apparent movement choice situa—
tions has been known for a long time. DeSilva (1926, p.
491—494) found that, although most Ss preferred counter—
clockwise movement, some exhibited a bias for clockwise
movement. Even Von Schiller (1933, p. 192) reports that
Ss were influenced unequally by experimental variation of
form and color in a choice situation. He found "two types"
of Ss: Some saw movement more easily between stimuli of
equal color, others between equal form. Willey (1936)
undertook a larger study on individual differences in
directional variations of apparent movement, by offering
his Ss choice situations which were systematically varied.
He concludes:

Perceptions of the direction of apparent
movement have been found to be characterized by
wide individual differences.... This diversity
in individual observations indicates the extreme
complexity of the interaction of the basic deter—
minants of these perceptions. Lack of knowledge
of the physiological processes underlying perceptual

activity makes it impossible to explain these changes
in perception, except as they can be related to the

21

influence of hypothetical "biases," "factors" and

"tendencies." With the use of such concepts the

results of experimentation on directional variations

of apparent movement can be made more intelligible.

Individual biases may include both a vertical or

horizontal directional tendency and a rotational

(clockwise or counterclockwise) directional ten—

dency.... The relative strength of a perceptual

bias can be estimated by the resistance it offers

to a change in the perception of movement when the

conditions of stimulation are modified to favor

such a change. (Willey, 1936, p. 673)

Since the relative strength of perceptual bias is
not known for each biased S, we have to restrict our pre—
dictions to relatively unbiased Ss. It may be thought that
being unbiased may not mean exhibiting 100% perception of
total assimilation. Such factors as set (instructions)
may prompt S’to see directional connotations when practice
has not yet taken place. This set, however, should not be
strong enough to weaken completely the tendency to total
assimilation. Even if S is somewhat set to see differences
to begin with, he should not show a strong bias for one
direction or the other. A stricter definition of bias,

ruling out even the slightest preference of one side over

the other, is of course possible.

1.7: The hypgthesis. With these cautions in mind, the
hypothesis of this study can now be stated in a more suc—
cinct form:

(1) Unbiased Ss who, in a previous performance, looked
repeatedly at unilateral apparent movement while actively
reinforcing directionality with motor gestures and ver—

balization, will on a later performance, exposed to a

22

bilateral apparent movement pattern, increase the number of
reports with connotation in the direction of practice.
(2) The null hypothesis can be stated as follows:

The proportion of Ss increasing the number of reports
with specific directional connotation (left, right), and
Ss responding in any other way, will not be significantly
different between the unbiased experimental groups with
unilateral practice and the unbiased control group without
such practice. Also, the mean pretest to posttest change
in reports with specific directional connotation (left,
right) will not be significantly different between experi-

mental and control groups.

2. Method

2.1: Apparatus. The experimental equipment, a modification
of that used by Tech and Ittelson (1956, p. 203f), consisted
of a viewing apparatus connected to a control apparatus,

a lever apparatus, and a dark alley.

(a) The viewing aoparatup was a rectangular box,
resting with its longer side on a support which in turn
was mounted on a table. The ipppp of the box, facing
toward S, consisted of a frame with a transparent mirror
(screen). The inside of the box contained three separate
compartments, 2.5 x 5 inches each, divided by tin parti-
tions. These partitions were fronted by diffusion glass.
Each of the compartments enclosed two 7.5—W filament bulbs.
The sockets for the bulbs were mounted on the pgpk of the
box, which could be lifted from the viewing unit in order
to inspect or exchange the light bulbs. The box could
also be opened on one side. From this side, by means of
guides, the diffusion glass on the inside and the trans-
parent mirror on the frontside were slid in. Between
diffusion glass and mirror there was room for a metal
slide holder. This slide holder could contain three 2.5
x 2.5 "transparencies" in a row. Each transparency was
thus centered in front of one of the three compartments.

By opening the viewing box laterally, and pulling the
23

24

slideholder halfway out, the transparencies could be changed.
From the back of the viewing unit wires led via a switch-
board into the timing unit, thus connecting the light
sockets to different timers.

(b) One part of the control apparatus was a timing
HEAL: a larger box holding a set of electronic interval
timers. These timers, produced by the Hunter Company, are
silenced and can be hand—operated by means of switches.
For the purpose of this experiment, three timers wired in

series were sufficient to provide accurate control over

(1) the ppguence in which the bulbs in the com-
partments of the viewing unit were lit;

(2) how long each pair of bulbs in one compartment
were lit (flash—duration);

(3) the interval between the lighting of each of
the three compartments (flash-interval).

The first of the three timers regulated the duration
of the first flash; the second regulated the interval be—
tween flashes; the third, the duration of the second flash.

Another part of the control unit was the ygriable
.L:§_sformer connected to a voltmeter. This transformer
regulated the voltage of the current entering the timing
unit and hence the brightness of the bulbs in the viewing
unit. A constant voltage was set by turning a dial until
the desired reading (100 volts) on the voltmeter was
obtained.

(0) For the experimental groups of the present

investigation, a lever apparatus was used. It consisted of

a rectangular frame of approximately 3 x 2 feet, put on

25

four legs of approximately two feet in height (like a table
without table top). S was to sit in front of one of the
longer sides of the frame. Two levers were fastened on
top of the frame, on the side opposite S. These levers
reached approximately five inches over the front of the
frame and could thus be comfortably grasped. Rotating in
their fulcrum on the side opposite S, the levers could be
slid over the top of the frontal side of the frame, one to
the right, the other to the left. At both ends, on the
front of the frame approximately 18 inches from the center,
contact switches were fastened. These switches stopped
the movement of the lever. At the same time an electric
circuit was closed. As a result, a small pilot light at
S's recording desk was lit, signalling that the lever
movement had been executed. A door spring fastened to each
lever made it snap back to its original position.

(d) Another part of the experimental equipment was

a dark alley of approximately ten feet in length and four

 

feet in width. The control apparatus and the viewing unit
were placed on a table at one end of the alley, the lever
apparatus and a chair for S_at the other end. The alley
was constructed by suspending, along both sides, a curtain
of black fabric. The curtain hung by means of curtain
rings from a tightened wire. Access to the alley was ob-
tained by opening the curtain at either end. This allowed
Ss to reach their seat at one end, and.§ to change the

transparencies in the viewing box at the other end of the

Fig. 1 Dark alley, viewing box, and lever apparatus.

“

 

 

 

2?

alley. Experimental Ss were seated in front of the lever
apparatus, at one end of the dark alley, in a BREAK fixed
to the wall. The chair was provided with a head rest in
order to keep S's gaze approximately directed toward the
center of the viewing unit (screen). The fixed position
of the chair kept S at a distance of seven feet from the
screen. The timing unit was placed at the end of the alley
opposite S's chair. The timer switches faced S, sitting
outside of the alley at his :ecording‘depk. This desk was
provided with a lamp, containing a red bulb. The lamp
illuminated only the immediate area of the recording sheet

on the table.

2.2: Stimuli. The stimuli were slides, one featuring a
set of two, and another a set of three, circular yellow
lights. They were prepared by punching equidistant circular
holes in black cardboard of 2—1/2 x 7—1/2 inches. The
holes, 1/2 inch in diameter, were lined up horizontally
with a distance of 1-5/8 inches between them, and were
centered over the three (or two) bulb compartments in the
viewing box. In order to reduce brightness, a yellow
Ozalit filter was fixed behind each hole. This filter
lowered the brightness by ca 7%.

The brightness of the three stimuli was assessed
by a Macbeth illuminometer. The circle centered over the
middle bulb compartment measured 2.03 c/ft2 without filter

and 1.38 c/ft2 with filter. The circles centered over the

28

left and the right bulb compartments measured 1.76 c/ft2
without, and 1.20 c/ft2 with filter. This brightness level
was maintained by the variable transformer set at 100

volts. By lighting the two bulbs in the center compartment,
the central yellow light was flashed first with a duration
of 250 ms. After an interval of 12 ms, by lighting the
right and the left compartment at the same time and for

the same duration, the two outside lights were flashed for
250 ms. These durations and this interval were used by

Von Schiller with Ss who had difficulty in perceiving
apparent movement under the normal conditions of his experi-
ment. Pilot work had shown that more Ss reported apparent

movement with short flash durations and interval.

2.3: Subjects. One hundred and ten male and female Ss
participated in the experiment. They were mostly college
freshmen, students in introductory psychology courses at
Michigan State University during the summer of 1960.

Most Ss were obtained by passing out signup sheets in classes
enabling S8 to make appointments at their convenience.

By varying experimental and control conditions frequently,
Ss were randomly assigned to these conditions. None of
these Ss had served before in a similar experiment nor did
any of them have sophistication in experimental psychology.
Ss can thus be classified as "naive" observers. For various
reasons to be considered later, many Ss had to be discarded.
The data of 63 out of 110 Ss are considered in the present

study.

29

2.4: General pgocedure. The design of the experiment was
of the "before—after" variety. It consisted of a prelim—
inary phase, a screening phase, a pretest, a practice period,
a posttest, a posttest check, a posttest interview and a
questionnaire. The whole procedure lasted approximately

30 to 35 minutes: circa five—ten minutes preliminaries,
screening phase and pretest, circa five minutes practice,

and circa five—ten minutes posttest and posttest check and
circa five—ten minutes posttest interview and questionnaire.
This design imposed itself after considering that the
dependent variable was not perception of apparent movement

as such. It was the number of times "qualitative connota—
tions" in perceived apparent movement were reperted, or

more precisely, the number of times S reported a difference
in the apparent movement pattern given by the Von Schiller
conditions. This difference was expressed in terms of a
qualitative preference. A typical verbal response would

be: "The movement to the right is quicker." The like—
lihood that at least some apparent movement would be per—
ceived was maximized by choosing appropriately short,
standard flash durations and intervals, which would pre—
sumably be successful with a majority of Ss. In addition,
the initial instructions were aimed at setting S for per—
ception of apparent movement. Many investigators, as men—
tioned before, from Wertheimer to De Silva to Dimmick and
Kelly have emphasized the importance of "set," "instruction,"

and "suggestion" in maximizing perception of apparent move—

30

However, the introduction of such a set increased
the possibility that results favorable to the hypothesis
might be due to unintended cuas presented.by S or to the
compliance of S who could "see through" the experimental
design. In order to minimize this effect, the practice
period was presented to Sbas an experiment in its own right,
using the same apparatus but unrelated to the "differen—
tiation hypothesis." It became an "experiment in reaction
time and coordination of response." The pretest which
was to be used to establish a baseline of perceptual per—
formance, had to become a "rehearsal" of an experiment in
"seeing small differences in apparent movement." The post—
test had to become the latter experiment itself. Since
the "differentiation experiment" was preceded by a rehear—
sal, the "reaction time experiment," too, had to have its
rehearsal. If‘S did not see through the design, he per—
ceived the situation as consisting of the following phases:

(a) General introduction

(b) Introduction to the phenomenon of apparent

movement

(c) Rehearsal of experiment 1

(d) Rehearsal of experiment 2

(e) Experiment 1

(f Experiment 2

(g) General interview
More specifically, the procedure of the experiment was as
follows:

(a) In the_pre1iminary_phase, S was introduced to

the experimental room, seated, and told in which direction

to look. Then the experimental room was completely darkened

31

except for the red recording light at st desk. While S
was dark-adapting, S recorded name and sex of S and assigned
S to one of the two experimental or to the control con-
ditions. S was made to feel as comfortable and relaxed as
possible under the circumstances.

(b) The screening ppggp was aimed at establishing
whether S saw apparent movement at all. Only Ss with an
apparent movement threshold compatible with the standard
conditions could be valuably retained for practice. After
the preliminary instructions, Ss were told that they were
participating in two separate experiments, but that the
same apparatus was used in both. This apparatus was said
to produce apparent movement stimuli. Ss were asked whether
they knew what apparent movement was. If Ss did not know
the expression "apparent movement," the phenomenon was
carefully explained. After this explanation, the first
trial of the screening phase was presented. The center
light in the viewing box was flashed and followed by a
flash to the right (a—bg sequence). S was asked: "What
did you see?" After S's report, the center light was
flashed again, this time followed by a flash to the left
(b1 —a sequence). S was asked again to report on his
perception. Finally, the last trial of the screening phase
was shown: The center light was flashed first and was
followed by the simultaneous flashing of lights to the

right and the left (bl -ar b2).

(c) The screening phase was divided from the pre—

test by a "rehearsal" of the training phase. The experi—

 

mental group was led to believe that the training phase

was an "experiment in reaction time and coordination."

The "rehearsal" consisted of making a quick lever movement
in synchronism with a perceived apparent movement. Ss in
experimental subgroup I were given one trial with an appar—
ent movement stimulus to the right (the same as the first
trial of the screening phase). They had to move the right
hand lever to the right. Ss in experimental subgroup II
were given one trial with an apparent movement stimulus to
the left (the same as the second trial of the screening
phase). They had to move the left hand lever to the left.
Ss in both experimental subgroups were told that in the
"real" experiment they would have to say a phrase indicating
the direction of their movement. Ss in the control group
were told they were participating in an "experiment on time
perception." They were instructed to pay attention to

two "clicks" of a stopwatch, one of which signaled the
beginning and the other the end of a time interval. Then
they reported their estimate of the length Of that interval.
For all groups, the "rehearsal" consisted of one trial

only in order to minimize the possibility of a "practice
effect" on the subsequent pretest. The control group re—
ceived instructions and one trial of fifteen seconds between

"clicks" as a rehearsal of their time estimation experiment.

33

(d) The ppgpppp was masked as rehearsal of a second,
different experiment aimed at assessing S's "accuracy in
perceiving small differences in quality" of apparent move—
ment. The true function of the pretest was to obtain an
initial level of perceptual performance. By repeating the
same test later, after the practice phase, every change in
the level of performance could be attributed to the inter—
vening practice. The pretest was also used to separate Ss
without initial "directional bias" from S3 biased to the
left or right side. S was designated "biased" if he re—
ported in more than three out of nine trials a difference
in quality in one direction of the bilateral Von Schiller
condition. "Unbiased" Ss were expected to report bilateral
movement at least one—third of the time. The pretest con—
sisted of ten presentations of an identical stimulus con—
dition. The first trial served as "warm—up" trial, leaving
a total of nine trials for the test proper.

The instructions contained the hint that the pre—
sentations of the pretest might be different. In reality,
on each of the ten trials of the pretest, the light in the
center was flashed first for 250 milliseconds and was
followed after 12 milliseconds by the flashing of the two
lights, to the right and to the left, for 250 milliseconds
(bl -2— b2 sequence as in the third trial of the screening
phase). S reported after each trial whether he saw any
difference in quality between the left and right side of

the bilateral apparent movement. The trials were given

34

every three seconds, leaving enough time for S to respond
and S to record the response. Also, an interruption was
needed in order to avoid a "carry—over" of the preceding
perception to the next trial.

(e) After the pretest, each experimental S received
fifty directional ppactice trials. One experimental sub—
group practiced exclusively to the right, the other ex-
clusively to the left. A single trial consisted of S's
observing the viewing unit and moving the lever to the
right (left) as soon as an apparent movement to the right
(left) was perceived. At the same time S had to pronounce
the phrase "to the right" ("to the left"). ‘S was thus to
make three different responses at the same time, a per-
ceptual response, a motor response, and a vocal response.
Hence, the representation of the training phase as an
"experiment in coordination of responses and reaction
time" was probably convincing to S. Unilateral apparent
movement was produced by flashing only two of the three
lights, the central one first, followed by the light to
the right for one group, the light to the left for the
other group. When moving the lever in synchronism with
his perception in the given direction, S had to hit the
contact switch at the end of either front of the lever-
apparatus. This was said to "register" S's reaction time.
In reality, the contact switch lit up a pilot light at Efs
desk and indicated that the lever was pushed by each S as

quickly as possible and for a standard distance of about

35

15 inches from the center of the apparatus. Ss were en—
couraged to move the lever vigorously and to pronounce
the phrase loudly and clearly. After each lever movement,
they had to let the lever snap back into the original
position and to initiate each lever movement by grasping
the lever anew. The practice period lasted about five
minutes.

The equivalent, in the control group, of the practice
period, was an "experiment in time estimation." After the
instructions and the aforementioned preliminary trial of
15 seconds, Ss were now presented with four time intervals
of one minute each. The beginning and end of the intervals
was marked by the "click" of a stopwatch. S had to give
his estimate of the length of the interval immediately
after the second "click" marked the end of the interval.

By its very nature as a "time experiment," it was easy
to equate the duration of the control treatment with the
average duration of the experimental practice phase.

(f) The posttest was an exact replication of the pre-
test. S was given the same instructions, and the same
ten trials of the bilateral Von Schiller condition (spaced
three seconds apart) were presented again.

(g) The pgsttest integview was held in order to

 

obtain further information from S on how "real" his ex—
perience of qualitative differences within the Von Schiller

pattern had been and whether he continued seeing apparent

36

movement during the tests. The experimental room was lighted
completely and the interview was initiated by S asking S
whether the "first or the seCond experiment" had been "more
difficult." The purpose of the interview was to obtain
more information on the nature of S's experience following
practice. After consulting the posttest record, a typical
question S could have asked was: "You reported on one of
the last trials that the movement toward the right side

was faster. Could you tell me a little more about how

that looked?" If the description of S's experience was

not enlightening enough, drawings were sometimes elicited
during the questionnaire period. The last question of the
interview was aimed at assessing whether the apparent move—
ment experience as such continued throughout the experiment.
S may have phrased his question: "Did you continue seeing
apparent movement all the time?" Written records were

kept on S's responses to the interview questions.

(h) A ghggk pp the posttest was given in order to
see how persistent any change in perception due to practice
was. At the end of the interview, §_remarked casually,

"By the way, would you mind looking at one more of the
trials from the second experiment?" Then the light was
turned off and S was quickly dark-adapted-and presented
with the Von Schiller situation again. After the pre—
sentation of the trial, S asked "What did you see?" and

recorded the response. Then the light was turned on again.

37

(i) The questionnaire was given to S after the
posttest check. It was aimed at finding out whether S had
seen through the experimental design and what his level of
awareness and attitude was with respect to the whole task.
The questionnaire was of the sentence completion variety,
but provided space for spontaneous comments of S, The
instructions read: "Please complete the following sentences
and add as much as you can." The first sentence started:
"The purpose of this experiment was..." The second sentence
was: "My evidence for this is..." The third sentence
read: "My attitude toward this experiment was..." ‘S was
completely alone when filling out the questionnaire.

Either S left the experimental room while S was working on
the questionnaire, or S was instructed to fill out the

form outside of the experimental room while S continued the
experiment with the next S. This was deemed to be a neces-
sary control in order to avoid any suggestive influence by

E.

2.5: ‘gnrlipi f Data. In before—after designs, the

 

pretest has the function of setting a baseline to which an
eventual change (due to experimental treatment) can be
compared. It was assumed that training might affect
initially biased.Ss (preferring one or the other direction)
differently from unbiased Ss. S was designated unbiased if,
given the Von Schiller situation, he made his reports in

the pretest by guessing on a chance basis and/or if he

38

reported no directional connotation most of the time.

That S might report some connotations in the pretest should
be expected since he was instructed to look for differences.
In a lenient definition of bias Ss with a very slight bias
might be included among the unbiased Ss, attributing the
slight bias to chance rather than to preference. An un-
biased S would then be one who in nine trials reported
connotations to the left three times or less, while re—
porting connotations to the right three times or less. A
typical series of reports from such an unbiases S might
look like this:

"no difference"

"no difference"

"quicker to the right"

"no difference"

"faster to the left"

"faster to the left"

"no difference"

"faster to the right"
"quicker to the left"

\OCCDflOKJ‘k-(Z'WNH
o o o o o o o o o

S would have to produce on the pretest one of the results
presented in Table 1 in order to be included among the
"unbiased." Since the same pretest was given to the control
group, the proportion of Ss with no bias and Ss with a

slight bias should not initially differ significantly for

the experimental and control groups. If the result of a

chi square test between the two groups is not significant,

it may be assumed that they are random samples from the

same population. Conversely, if training were effective,

one would expect a significant difference between experimental

Ss and their controls on the posttest.

39

Ss were classified as "biased" if they reported
connotation in a specific direction more frequently than
they would if guessing at random. A typical record of
such a S might look like this:

"no difference"

"no difference"

"mire movement to the right"
"mire movement to the right"
"no difference"

"more movement to the right"
"quicker to the left"

"more movement to the right"
"faster to the right"

\OGD\10\U\‘P\A)Ni-‘
o o a o o o o o 0

Such a bias, whatever its specific nature may be, presumably
interacts as an organismic variable with practice. Thus
practice might on the whole decrease bias. Or practice in
the direction opposite to the bias might conceivably increase
the bias ("boomerang effect"). It might also decrease
the bias (but not increase connotations in the trained
direction); or it might provoke some kind of "reversal
effect," replacing the original bias with another equally,
or nearly equally, as strong bias. All these hypothetical
reactions would make a biased S a bad candidate for this
experiment. Fortunately, these Ss can be detected through
the pretest. Unfortunately, however, the proportion of
biased and unbiased Ss in our population is unknown. The
bias itself may come from different sources such as eyedness,
handedness, having learned to read from left to right
(or right to left), etc. Short flash durations and a short
interval may on one hand enhance perception of apparent

movement as such; on the other hand they may increase the

40

Table 1

Number of connotations to be reported on pretest by relatively

unbiased.Ss

 

 

 

 

 

Connotations
degree of bias
left none right
no bias ’ 3 3 3
l 7 - l
9 ' O
slight‘bias ‘ ' — 2* {‘4 — i 3 _
1 5 3
0 | 6 l 3
I 3 4 2
3 5 l
3 6 0
O 8 l
l 8 0
2 7 O
O ’ 7 2

 

 

 

 

41

proportion of biased Ss by making the stimulus pattern
more ephemeral and unstable (and thus open to S's prefer—
ential interpretations).

Since the proportion of biased and unbiased Ss cannot
be known in advance, it is difficult to specify a particular
statistical test before the data are collected. The nature
of the data suggests at first look a nonparametric technique.
If a large enough group of unbiased Ss were obtained, then
the posttest data could be subjected to a chi square test
for independent samples (experimental vs. control group).

If the number of unbiased.Ss is small, Fisher's exact
probability test would be appropriate. As a control for
the interpretation of the results the number of unbiased
Ss must be divided into three different treatment groups
(left practice, right practice, no practice). This makes
the available N so small that, in the chi square tests,
Yates's correction for continuity becomes necessary.

On the other hand, computing a difference score
based on the number of responses with directional connotation
before and after practice, would allow for a p test (pro—
vided the distribution of the difference scores is not
seriously skewed).

To assess whether practice had any effect in terms
of increase in directional connotation, the test would be

the following:

1,2

Initially unbiased experimental Ss who increase the
number of reports with connotation on the posttest are
tested against the unbiased controls who increase the num—
ber of reports with directional connotation on the posttest.
A difference score is computed for each S by subtracting
the sum of directional connotations (right plus left) on
the pretest from the sum of directional connotations on the
posttest. If the difference is positive, an increase has
taken place. If the difference is negative, the number of
scores with directional connotation has decreased. If the
difference is zero, S has not changed the total number of
scores with directional connotation (although he may have
regrouped the scores as to their specific direction, left
or right).

A chi square test may be used to test the proportion
of the S3 with positive difference scores Iand‘Ss with
negative and/or zero scores in the experimental group against
the proportion of Ss in the control group with positive
scores“ and Ss with negative and/or zero scores.

The difference between the mean difference scoreS_
of the experimental and the control group S may be tested
for significance with the more powerful 3 test.

With either technique, results significant at the
alpha = .05 level should be interpreted as supporting the
hypothesis.

The following two tests would be necessary for the

acceptance of the major hypothesis:

43

(1) Initially unbiased right trainees who increased
the number of reports with directional connotations to
the right after practice are compared with controls who
increased, without practice, reports with directional
connotations to the right.

(2) Initially unbiased left trainees who increased
the number of reports with directional connotations to
the left after practice are compared with controls who
increased, without practice, reports with directional
connotations to the left.

"Increase" is defined as any deviation from the
baseline established in the pretest. An increase score is
computed for each S,by first computing a base score for
the pretest, from which a posttest score is subtracted.
For Ss with right practice, the pretest score is obtained
by subtracting the number of initial reports with left-
side connotation from the number of reports with right—
side connotations. A negative difference would indicate
a slight left bias, a positive difference a slight right
bias, a zero difference complete absence of bias. For S8
with left-side practice, the pretest score is obtained by
subtracting right—side responses from left-side responses.

In the same way a posttest score is computed for
Se in both groups. Then pre— and posttest scores are
compared. If the posttest score is a higher numerical
value than the pretest score, an increase of directional

connotation in a specific direction is said to have taken

44

place. This increase can be expressed as a change score
from which an average change can be computed. This average
change score may be compared with the average change in

the same direction in the control group (3 test). A

less powerful test would be a chi square comparison between
number of experimental Ss changing in the specific direction
of the practice vs. control Ss changing in the direction

of practice.

With both techniques, results significant at the
alpha = .05 level can be interpreted as support for the
basic hypothesis and for the explanation that practicing
a specific direction (viz., left or right) increases di—

rectional connotation in that specific direction.

3.1: Sgggpg‘gpd Unbiased _§. The pretest was devised,

in part, to single out Ss who, given the standard experi—
mental conditions, were directionally biased to begin with.
Ss were classified as unbiased when the distribution of
their preferences appeared to be random or when they did
not exhibit marked preference at all. Out of 63 male and
female Ss who saw apparent movement, only 40 were unbiased.
This relatively low percentage includes five _s who did
not see apparent movement during the screening phase but
said in the posttest interview that they had seen apparent
movement throughout the experiment. The other 23 Ss, who
all saw apparent movement during the screening phase,

exhibited a strong bias in one direction or the other

during pretest.

3.2: Pretest pp ppsttest change 1p diregtiongl connooation.

 

The results of the experiment are reported in Table 2.
Of the 40 unbiased Ss, 23 were assigned to the experimental
condition and 17 to the control condition. Out of 23
unbiased experimental'Ss, 17 increased, two decreased, and

three did not change the number of reports with directional

connotation from pretest to posttest.

45

46

Out of 17 unbiased contrg1_Ss, six increased, five
decreased and six did not change the number of reports
with directional connotation from pretest to posttest.

Out of 15 biased experimental Ss, three increased,
seven decreased, and three did not change the number of
reports with directional connotation from pretest to post—
test.

Out of eight piased contro1_Ss, three increased,

 

four decreased and one did not change the number of reports
with directional connotation from pretest to posttest.

Inspection of the data suggests that there is a
difference in pretest to posttest change between unbiased
and biased Ss. Whereas unbiased.Ss increase their number
of directional connotations, biased Ss seem to decrease or
not change the total number of directional reports. This
hypothesis was tested by applying a chi square test to the
data presented in Table 2 (Ss who increased.vs. Ss who did
not change or decreased.) The Chi square results were
significant beyond the .01 level. The hypothesis that
there was a difference in effectiveness of the experimental
treatment between unbiased and biased Ss is strongly supported.

The data also suggest that unilateral practice was
effective in increasing directional connotation for unbiased
Ss. A chi square test applied to the change data for un—
biased.Ss in Table 2 resulted in significance beyond the .05
level. The hypothesis that active unilateral practice with
apparent movement increases the directional connotation in
perception of bilateral apparent movement patterns gains
support.

Mean differnce scores between pretest and posttest

were computed for the unbiased experimental and control

groups. The significance of the difference of the two

47

Table 2
Pretest to posttest change in number of reports with directional

connotation for unbiased and biased Ss

 

 

 

groups N incr. deer. no change
unbiased 4O 23 7 10

1
exp. 23 ' 17 l 2 4
contP. 17 6 I 5 I 6

i
biased 23 5 12 - 6
exp. 15 , 3 7 5
cont?. 8 l 2 5 l

 

 

 

 

 

 

48

means was then tested by a 3 test. The p was significant
beyond the .05 level and the hypothesis that the mean change
score in directional connotation is a result of practice
with unilateral apparent movement receives further support.

The results of statistical tests are presented in
Table 3.

Table 4 presents central tendencies and variability
in the number of reports with directional connotation from
pretest to posttest. The data suggest that the variability
among Ss in the number of reports with directional conno—

tation increased in all unbiased and biased groups alike.

3.3: Pretest pp pppptest chapge with pppcific directional

connotation. What is there about practice that made ex—

perimental Ss increase the number of reports with directional
connotation more often and on the average more than their
controls? Was it the specific direction of unilateral
practice (to the right or to the left) as predicted in the
major hypothesis of this study? The increase of the number
of reports with specific directional connotation is shown
in Table 5.

Out of nine unbiased Ss with practice pp_ppp_plgpp,
two increased the number of reports with connotation in the

direction they had practiced, whereas five changed in the

direction opposite to that practiced. One S, while decreasing
the total number of directional connotation reports, neverthe—
less increased the number of reports with connotations in the

direction opposite to his practice. One S did not change at all.

Summary of chi square

49

Table 3
and p tests for differences between

groups

 

groups I

chi
square

df

M1

3 SDlv

SDZ

df

‘

 

-‘-

unbiased
vs. biasedi

 

1

7-563I*.1

v
D

i

1.48

2.62

1.71

 

3.878*1 61

 

unbiased I

exp. vs.
controls

4.490*

 

1.98

;
l

2 629*!) 38

--.._J-.1_.

 

 

biased
exp. vs.
controls

.170

 

.038

 

 

unbiased
rt. pract.
vs. lf.
pract.

—_._

1.483

. 'iI

p.699\ 21

 

Ehbiased—_‘

rt. pract.
vs.
controls

I
2.056

2.46

.901. 24

 

unbiased
1f. pract.
vs.
controls

3.648

 

“"

1.78

L.l49

 

 

totally
unbiased
rt. pract.
vs. 1f.

 

pract.

 

 

 

2.00

—l.00

 

 

.82

 

 

 

 

5.676%

 

-.~—-

 

*-

*-

*-

‘1'—

Significant at the .05 level.

Significant at the .01 level.

Central tendencies and variability of the number of

directional connotations reported by unbiased

 

50

Table 4

and biased Ss

 

 

 

 

 

 

 

group;--~Ifi-;;I—-__pretest #— f 'posttest
Q2 Q l 02 5 Q

.
unbiased 40 l 1.94 41.07 3.50: l 2.00
pract. rt. 9 i 2.33 i1.33 4.50 1.73
pract. 1f. 14 1.50 .98 4.50‘ I 1.44
experim. 323 1.92 1.11 f 4.40 1.48
control 17 2.00 I .81 i 2.50 ' 1.61
biased 23 I 6.00 1.11 , 5.67, 1.55
pract. rt. 7 | 7.00 11.52 i 6.75' l 2.19
pract. 1f.I 8 6.50 {1.17 I 6.00' E 1.25
experim. '15 I 6.67 [1.40 l 6.33 1.47
control 8 | 5.75 .63 4.50 ; 1.08

 

 

 

 

 

 

51

Table 5

Increase of number of reports with specific directional

 

 

 

 

 

 

 

 

 

connotation
groups N } incr. rt. incr. 1f. 1 other
1' '

-__. 1--,-- _ ______ __ ._ .__
rnbiased 4O 13 14 ‘ l3
pract. rt. 9 2 6 l
pract. 1f. 14 8 I 3 3
experim. 23 10 I 9 4
control I 17 3 i 5 9

I I I
biased I 23 7 8 I 8
pract. rt. 7 i 2 3 2
pract. 1f. 8 2 I 2 4
experim. 15 'I 4 I 5 6
control I 8 3 3 2

I I

 

 

 

 

 

52

Out of 14 ppbiased Ss with ppagplce to the left,

4'- ~‘- ”

 

three increased the number of reports with connotation in
the direction they had practiced, whereas eight changed in
the direction opposite to that practiced. Two Ss increased
the number of reports with directional connotation equally
in both directions. One S did not change at all.

Out of 17 unbiased Ss who received.pp directional
practice, three increased the number of reports with
directional connotation to the right, five increased to the
left, five increased the number of reports without direc—
tional connotation, and four did not change at all.

Out of seven biased Ss with practigg pp the right,
one increased directional connotation in the direction
practiced; one increased the number of reports with con—
notation in the direction of practice, but also decreased
the total number of reports with directional connotation.
Three Ss increased the number of reports to the direction
opposite practice. Two Ss increased the number of reports
without directional connotation (one of them at the expense
of both right and left connotations, the other at the
expense of right connotation only).

Out of eight biaseg Ss with pracpice t the left,

 

two increased the number of reports with directional con—
notation in the direction practices. Two Ss increased
directional connotation opposite to the direction practiced.
Three Ss increased the number of reports without directional

connotation (one of them at the expense of both right and

53

left connotations, the other at the expense of right
connotation only). One S did not change at all.

Out of eight biased Ss who received.pp practice,
one increased the number of reports with directional con—
notation to the right; two increased the number of reports
with connotation to the right and increased also the num—
ber of reports without directional connotation (both at the
expense of left connotation). Two Ss increased the number
of reports with directional connotation to the left; one
increased to the left but also increased the number of
reports without directional connotation (both at the ex—
pense of right connotations). Two increased the number of
reports without directional connotation (one at the expense
of right connotation, the other at the expense of left
connotation).

The data suggest that practicing a specific direction
did not increase the number of reports with directional
connotation in that direction. There seems to be rather a
tendency to the contrary in the unbiased experimental group;
practicing a specific direction apparently increases the
number of reports with directional connotation in the
direction opposite of practice.

To test whether practicing a specific direction was
effective in producing any increase in reports with specific
directional connotation (be it in agreement with or opposite
to practice) the mean increase in specific directionality

for unbiased Ss with left practice was compared with the

54

mean increase in specific directionality for unbiased Ss
with right practice. A p for the difference of means was
computed and found not significant.

Then the proportion of Ss with right practice who
increased the number of reports opposite to the direction
of their practice was compared with the proportion of Ss
in the control group who increased the number of reports
with left connotation. The computed chi square was not
significant. When, however, the proportion of Ss with
left practice increasing connotations opposite to their
practice was compared with the proportion of control Ss
who increased reports with connotations to the right, the
computed chi square just failed to reach the .05 level of
significance. Such a result might possibly indicate a
trend in the left trained group to react in the direction
opposite to training.

Could the reason for the fact that the trend does
not reach significance be found in the lenient definition
of bias? Even the slightest initial bias, that is, any
net excess of directional connotation in one direction
or the other, might in fact interact with training.

In order to test this after the fact hypothesis, a
3 test was applied to the mean difference scores of the
totally unbiased.Ss. The number of totally unbiased Ss was
very small. The mean increase of reports with left con-
notations for the three unbiased Ss with practice to the

right was +2, while it was —1 for the three unbiased Ss

55

with practice to the left. A 1 for the difference of these
two means is significant well beyond the .05 level. For
the few totally unbiased.Ss in this experiment, practice

in a specific direction did have an effect in the opposite
direction of training. Given the small N, however, this
result should be interpreted with caution.

How did specific directional practice affect slightly
biased and strongly biased S8? The distribution of their
reaction to specific training is shown in Table 6.

In the group with slight bias, five Ss increased in
the direction of their practice after they had practiced
the direction opposite to their initial bias; only one S
behaved to the contrary. Ss who practiced the direction
of their initial bias increased their connotations in the
direction opposite to what they had practiced (no S behaving
to the contrary). Under both training conditions, a total
of three Ss did not change at all.

Ruling out the three Ss who did not change, the
hypothesis that such a distribution of cases could happen
by chance was tested. The computed chi square, corr., was
significant beyond the .01 level. One can have confidence
in the after the fact hypothesis that for slightly biased
Ss practice was effective in increasing directional conno-
tation in the direction of practice when it was against
initial bias, but when it reinforced initial bias, Ss changed

in the direction opposite to practice.

56

Table 6

Interaction of practice and bias

 

practice and bias

Slight bias

Tractice against
bias

Practice in
direction of
bias

'~

‘—

increase in
direction of
practice

L_,

increase in
direction
opposite
practice

no increase

 

 

inrong bias

Fractice against
bias

Tractice in
direction of
bias

 

 

 

 

 

 

 

57

The strongly biased group exhibits the same tendency,
although the computed chi square, corr., is not significant.
Combining both biased groups, chi square, corr., is sig—
nificant far beyond the .01 level. Any degree of bias then
seems to interact with training. Biased Ss practicing in
the direction of their initial bias increase connotation in
the direction opposite to their practice, while Ss practicing
the direction opposite their bias increase connotation in

the direction of practice.

3.4: ‘Ppggggp pp posttest decrease p: bias. Inspection of

 

the control group data indicates that, although Ss did not
practice any direction, they changed from pretest to post—

test. This change could be described as "decrease in

bias." While generally decreasing the total number of
reports with directional connotation, these control Ss
increased the number of reports with connotation opposite
the direction of their initial bias, mostly at the expense
of the number of biased responses. Often these Ss actually
reversed their initial bias, or became "unbiased," reporting
an equal number of connotations in both directions.

Any net excess of reports in the direction opposite
the initial bias on the posttest may be defined as "decrease
of bias." In the control group with slight bias, nine Ss
exhibited decrease, two increased in bias, and one did not‘
change. A sign test applied to these data is significant
beyond the .05 level. In the control group with strong
initial bias, all eight Ss decreased in bias. A sign test

Eapplied to these data is significant beyond the .01 level.

58

A sign test applied to the pooled control groups with
slight and strong bias is significant beyond the .001 level.

Did this decrease in initial bias also occur in the
experimental groups which increased their total number of
reports with directional connotations? Thirteen of the
slightly biased experimental Ss decreased in bias, one S
increased and three did not change the number of initially
biased reports. A sign test applied to these data is
significant at the .001 level.

In the strongly biased experimental group, ten‘Ss
decreased in bias, three Ss increased in bias and two did
not change. A sign test applied to these data is significant
beyond the .01 level. If the sign test is applied to the
posttest data of the pooled experimental groups with slight
and strong initial bias, the outcome is significant beyond
the .001 level.

What happened to the initial bias in experimental
and control groups with slight and strong bias? Out of
the total of 52 initially biased Ss, 20 reversed their
bias in the opposite direction, 13 became unbiased, re-
porting an equal number of connotations in both directions,
eight simply decreased bias (five of these were Ss with
particularly strong initial bias), five Ss did not change,
and six Ss increased their initial bias. The tendency of
most Ss to reverse the initial bias is equally strong in
experimental and control groups; the tendency to become

unbiased is the second most frequent of both groups.

59

3.5: Hompgeneipyjp£_the gpopps. One of the questions one
might ask in looking at the pretest data is whether the
assumption is tenable that the "unbiased" experimental and
control groups were selected at random from the same popu-
lation. If they were, in fact, selected from the same
population, then any difference between experimental and
control Ss from pretest to posttest could be interpreted
as an effect of the experimental treatment. The hypothesis
that all Ss who underwent the pretest were from the same
population can be tested at least in a limited sense. We
could, for instance, single out a common characteristic
among unbiased experimental and control groups which gives
some assurance that they were drawn from the same popu-
lation. Such a characteristic could be that the groups do
not exhibit a significant difference in the proportion of
'Ss with no bias and.Ss with slight bias. We would then
test the hypothesis "experimental and control groups differ
with respect to their proportion of unbiased and slightly
biased.Ss in the pretest" against the null hypothesis
"experimental and control groups do not differ with respect
to their proportion of unbiased and slightly biased Ss."

A chi square test was first applied to the pretest
data of both experimental groups. Chi square, corr., was
not significant. The null hypothesis cannot be rejected
for the two experimental groups. The pretest results of
the two experimental groups were then pooled and another

chi square test was applied to the pooled data of the two

60

experimental groups vs. the control group data. Chi square
corr., was not significant. The null hypothesis that
"experimental and control groups do not differ with respect
to their proportion of unbiased and slightly biased Ss"

cannot be rejected.

3.6: Ss seeipg through the expepimental design. S's level
of awareness was one of the problems of experimental control.
If Ss were aware of E's hypothesis, their verbal report
could be viewed as a result of compliance or opposition
rather than an indication of their perception. It turned
out, however, that out of the 63 Ss only one gave evidence,
in the questionnaire response, that she did understand the
hypothesis of the experiment. This S wrote: The purpose
of these experiments was:
to see if the idea of the direction 'right' would
carry over (by continuous repetition) to an experi-
ment which obviously didn't contain these movements.
My evidence for this is the fact that the lever
experiment came before the direction one. My
attitude toward this experiment was very neutral;
it seemed confusing.
Six other Ss expressed curiosity, doubt or speculations as
to whether S's instructions actually presented the real
purpose of the experiment. One typical control S wrote
that the purpose of the experiment was:
to determine perception of time and small differences.
However, I do not know what or if there was an un-
derlying purpose, but I feel that there was. My
evidence for this is (1) tests (2) these tests
didn't seem to be lending[?] to just the purposed

conclusion. They seemed to be getting at something
deeper that the announced.

61

Another S wrote:
I did not get the full benefit of the second test
because I was not sure what was trying to be proved.
I definitely felt that the two tests were exact
opposites....
Another S, though accepting S's instructions as the "real
purpose" of the experiment, mentioned a "carry over" from
the first to the second experiment. 'S wrote: The purpose
of these experiments was:
to measure my ability to coordinate my verbal hand
reactions to the perception of stimuli by eye.
The second experiment was designed to measure my
perception of differences in a stimuli [stimulus]
with two reactions. (I felt a carry—over from the
first experiment to the last in the form of increased
awareness of the experimental material) after thought.
Another revealing comment as to the expectation.Ss may have
had, despite the instructions, was given by an.S who re-
ceived right side practice:
My attention was not completely on movement to the
right because my left hand became tense in antici-
pation of any movement to the left.
Most Ss accepted the "purpose" given in the experimental
instructions readily as the "real purpose" of the experiment.

It can be concluded that the "rigging" of the experiment

by the instructions was, on the whole, successful.

3.7: Attitude. Even if S did not see through the design

 

of the experiment, a neutral, positive, or negative attitude
toward the experiment could have resulted from the different
treatments. To control for such an influence, the last
sentence to be completed in the questionnaire asked for

gi's attitude toward the experiment. It was found that Ss

62

pronounced in their writing either a general attitude
toward the experiment as a whole, or used preferential
statements to express a differential attitude toward "both
experiments." Of the first group of 45 Ss, 29 Ss expressed
a positive attitude toward the experiment as a whole. One
typical S wrote: My attitude toward this experiment was
"excellent. I tried to cooperate to the best of my ability."
Sixteen Ss expressed attitudes other than positive. A few
Ss expressed an obviously negative attitude. Here is an
example: My attitude toward this experiment was:

that I thought it was quite difficult to detect any

differences + it is hard to be positive in such an

experiment.
Other Ss expressed an attitude of indifference as, for

example, this: My attitude toward this experiment was:

one of indifference since there wasn't much time
put into it.

"Ambiguous" was the classification applied to some mixed

comments as the following: My attitude toward the experi—

ment was:
one of nervousness: I thought it was very interesting.

Some Ss evaded stating their attitudes by writing comments

like the following: My attitude toward this experiment was:
I'll bet this research could be used by people in
advertising or public safety in determining the
effectiveness of using 'moving' lights to capture
the people's attention (e.g., for advertisements or
road signs.)

Out of 17 expressed preference, 15 were in favor of the

"first experiment" (the practice period, or, for control

.Ss the time estimation task). These Ss made comments like

63

this: My attitude toward this experiment was:
the first part was more fun; the more I concentrated,
the faster I could respond. During the second
experiment, I was uneasy because I could hardly
perceive the small differences and would feel nervous
while waiting for the lights to flash.
Only two Ss preferred the "second experiment" over the
first. One of these Ss, from the control group, commented:
My attitude toward this experiment was:
In the time lapse exp. there was a little pressure
and unrest because there was nothing concrete to
grasp, but the sound of the click of the stopwatch.
The second experiment was pleasant. There was the
blinking light to observe, wasn't left in the dark.
The question is whether different attitudes may have
been influenced by the particular treatments of the experi-
ment. Such a hypothesis can be tested against the null
hypothesis that the unbiased "experimental and control
groups were drawn at random from the same population with a
particular proportion of positive and other than positive
attitudes." A chi square test was applied to the attitu—
dinal data of all three unbiased groups. Chi square, corr.,
was significant beyond the .02 level. The null hypothesis
can be rejected; the hypothesis that the unbiased experi-
mental and control groups came from different populations,
as far as attitude is concerned, is supported. Did both
unbiased training groups come from populations with different
attitudes? Can failure to obtain any particular directional
trend in the right—practicing group be accounted for, in

part, by lack of a positive attitude in that group? Such

a hypothesis can be tested against the null hypothesis

64

that both training groups came from one population in which
positive and negative attitudes were represented in certain
proportions. If both groups were random samples from the
same population their proportion of positive and negative
attitudes should not differ significantly. Fisher's exact
probability was computed from the attitudinal data of

both experimental groups. The probability was .400, hence,
the null hypothesis cannot be rejected. The hypothesis that
the two groups stem from populations which differ fundamentally
in the proportion of S3 with positive attitude is not sup-
ported.

Was there a significant difference in attitude between
the left practicing and the control group? This hypothesis
can be tested against the null hypothesis that attitudes
did not significantly differ between left practicing and
control groups. A chi square test was applied to the
attitudinal data of the left practicing group and the
control group. Chi square, corr., was significant beyond
the .02 level. The null hypothesis must be rejected; left
trainees and control Ss seem to stem from different popu-
lations with respect to the proportion of positive vs.
nonpositive attitudes. (An overall difference in attitude
then seems to exist between left trainees and right trainees
on one hand and control Ss on the other hand; but only the
left practicing group contributes significantly to this

difference.

65

3.8: Posttesp check. If practice had produced some change
in S's perception of directional connotation, this change
could vary in its persistence over time. It could, for
example, be a short—lived set, dissipating quickly; or it
could be a more stable change, persisting over a certain
period of time. The posttest check, a one trial repetition
of the Von Schiller pattern, was given after the posttest
interview to obtain information on the persistence of the
posttest results. If experimental Ss reported in the
posttest check more directional connotations than control
Ss the increase measured in the posttest could be said to
be a stable one. Out of 23 unbiased experimental SS, 12
reported directional connotations on the posttest check,
whereas only three out of 17 unbiased control Ss reported
directional connotations. A chi square test was applied to
these data. Chi square, corr., just fails to reach the

.05 level of significance. The increase in directional
connotation measured in the posttest does not persist,
although there seems to be a tendency in that direction.

One trial only could be, of course, regarded as insufficient.

3.9: Ss reportipg apparent movement. In a preliminary
study, results other than those immediately connected with

time hypotheses are of interest. For instance, knowing

vduit proportion of Ss see apparent movement under the
descnribed experimental conditions may be of value for

fhncther investigation.

66

Out of the 110 male and female Ss, 86 reported
apparent movement during the screening phase. In the
posttest interview, 96 Ss said that they saw apparent
movement throughout the experiment. The capacity of S8
to see apparent movement seems to have increased slightly
from the screening phase to the posttest interview. Out
of the 13_Ss who reported in the posttest interview that
they seldom or never saw apparent movement, seven had not
been able to see apparent movement during the screening
test. Some Ss perhaps "learned" to see apparent movement
during the experiment as a result of repeated exposure to
apparent movement stimuli. On the other hand, only few Ss
reporting apparent movement during the screening phase,
said in the interview, that they were not convinced of

perceiving movement throughout the experiment.

3.10: Kinds pi apparent movement experience. When Ss

 

reported apparent movement, they did not always report
bilateral movement. Out of the total of llO‘Ss who par—
ticipated in the experiment, the data of only 63 Se could

be used. The reasons for discarding data were manifold.

Six Ss were eliminated because they were left-handers.

Some Ss presented scoring difficulties. But most eliminated
‘Ss did not report seeing apparent movement in the expected
way. Some discarded Ss reported movement in one direction
only. Many others reported movement in one direction,

followed by a second movement which began at the terminal

67

of the first movement and went in the opposite direction.
Generally this movement was described as "crossing the
center and ending up in another terminal." A typical
report was: "The movement went first from the center to
the right and then all the way over to the left." Very
few Ss reported a "lingering" of the terminals and/or of
the lights in the center. Some of these S3 with "deviant"
perceptual resolution of the Von Schiller pattern changed
during the experiment. Others maintained their mode of
resolution throughout. Others doubted whether they saw
movement at all. Often Ss who were discarded also exhibited

strong bias in the pretest.

3.11: Kinds p£_"connotations" reported. The instructions

 

called for a differentiation of qualities within an apparent
movement pattern. In order to avoid "forcing" the response,
no qualitative categories were suggested to S. ‘Ss were

thus thrust into a relatively uncomfortable situation,

not knowing what "qualities" they had to look for. Some

Ss attributed exclusiveness to the preferred direction.

They reported typically: "Moves to the right only." One

S reported: "I think I see more the right one than the

left one." Other reports attributed longer duration to

the preferred movement. Others reported the light in the

preferred direction "flashing back and forth." The pre—
ferred movement was also called "smoother." Sometimes a
report was simply of the order "more movement to the right."

This kind of criterion was similar to the one in which Ss

 

 

 

68

reported seeing "two bounces" on one side. Sometimes a
particular "brilliance" was attributed to movement in the
preferred direction. Ss reported "brighter movement" or

a "streak of light" in one of the two directions. More
often an "intensity criterion" was adopted. One movement
component was said to be "predominant," "stronger," "better,"
or "more positive" than the other. The criteria most
frequently adopted were those of distance—moved and speed.
The criterion of distance-moved was adopted by 28 out of

63 Ss at least once, the criterion of speed by 29 at least
once. Typical reports were: "The right one moved farther"
or "The one on the left went faster." Often the two criteria
were used together as in a phrase like: "the left one
moved further and quicker." Fortunately, most of the
unbiased Ss adopted the criterion of speed or distance-
moved in their reports. Only reports which revealed a
typical "stimulus error" were not scored. We define a
stimulus error as reports reflecting experience with the
light sources ("dots") rather than with the movement.

Thus phrases like: "The circle on the right looks bigger"
or "The left light goes on before the right one" were not
scored. When Ss reported connotations which were not
directional (e.g., "jumping up and down"), the report was
not scored. Similarly, contradictory reports were omitted
("The one on the left goes quicker and the one on the

right goes farther").

4. Discussion

4.1: Interaction pi practice gpd plgp. In this study a
group of Ss, relatively unbiased directionally,practiced
directionality by looking at unilateral apparent movement
while actively reinforcing the direction of that unilateral
movement. It was found that Ss with practice, when looking
again at bilateral apparent movement, increased their number
of reports with directional connotation significantly over
Ss without practice.

When Ss had a strong initial bias in any direction,
practice did not produce this effect.

Whereas the average frequency of reports with
directional connotation increased in unbiased experimental
Ss (and did not significantly increase in biased or control
lSs), the variability of the number of reports seemed to
increase in all groups, biased and unbiased.

Problems of interpretation of the results arise from
the fact that practicing unilateral apparent movement did
not increase the number of reports with connotations in
the direction which was practiced. While completely unbiased
Ss reacted in the direction opposite to their practice,
slightly and strongly biased Ss alike did not exhibit
significant counter reaction, although, at least among Ss

with left practice, there appears to be such a trend.

69

70

When Ss were grouped in terms of their initial bias,
it was found that practice did increase connotative responses
in the direction of practice only when it worked against
the direction of initial bias. When the initial bias was
reinforced by practice, Ss increased connotative responses
against the direction of their practice.

Matters are complicated further when it appears that
Ss without directional practice change from pretest to
posttest, not by increasing the total number of reports
with directional connotation, but by redistributing the
initial number in a direction opposite to their bias.

Most control Ss reversed their initial bias from pretest
to posttest, others simply decreased in directional bias.
The same tendency appears in experimental Ss.

Obviously, then, the effect of practice was super—
imposed on a second effect due to another factor (or factors)
within the experimental situation. This other effect can
best be observed in the control group, and consists in a
decrease and/or reversal in initial bias. What are the
factors which could be responsible for such a reaction?

It could be merely the repetition of the Von Schiller test,
or it could be the repetition of the instructions. Also,
'Ss who were set by the instructions to see directional
differences within the bilateral apparent movement pattern
may have reversed their bias simply because they expected
that the "rehearsal" could not be the same task as the

"real experiment." Such expectancy of a difference in

71

directionality between "rehearsal" and "experiment" could
have possibly interacted with specific directional practice.

When this practice reinforced S's hypothesis that the
"rehearsal" could not be the same as the "experiment,"
practice became effective. For instance, Ss who exhibited
right bias on the pretest would have reversed or decreased
this bias as a result of expectations aroused by some factor
in the experimental situation whether they practiced or
not. They would have increased the number of reports with
left connotations (probably at the expense of the initial
excess of report with right connotations). If these Se in
addition received left practice, their expectation would
be reinforced and they may have added further left responses
by diminishing the pool of responses without any directional
connotation.

If practice, however, would counteract the expec—
tation of bias reversal, by reinforcing the direction of
initial bias, one might predict a subtraction from the
effectiveness of practice since it interferes with expec—
tation. If practice was altogether a weak factor in this
eexperiment—_for reasons to be considered later--it might
even.have been overpowered by the expectation that the
'Nexperiment" had to be different from the "rehearsal."

131 this case practice would almost seem to "boomerang."
.All.that practice did then was to heighten generally the
eaxpectation of directional connotations. If this expec—

txation conformed to other directional expectations aroused

72

by factors other than practice, the result was an increase
of reports with connotations in the direction of practice.
Otherwise practice of a specific direction appeared to have
almost the opposite effect. This speculation seems to
coincide nicely with the results of Haggard and Babin (1948).
It will be remembered that they found active participation
(drawing arrows) in the direction of perceived autokinetic
movement was most effective in enhancing autokinetic move—
ment direction. But even active participation in the
direction opposite to the perceived autokinetic movement
made Ss perceive movement of greater distance, more fre—
quently and with more confidence.

Thus, in the present study, practice was signifi-
cantly effective in producing more directional connotation
whether it was done in the same or the opposite direction
of the "reversal effect" (produced by other factors in the

experimental situation).

4.2: Practice. It has already been suggested that practice
might have been an intrinsically weak factor in the ex—
perimental situation. This weakness could have several
causes. First of all, the number of practice trials may
not have been sufficiently large. In human conditioning
the number of necessary trials is extremely variable, de—
pending on the complexity of the conditioned response.

For example, in their experiment on conditioning of the

autokinetic phenomenon, Haggard and Babin (1948) used

73

five sessions with thirty trials each. In the present
study, the number of trials was kept at fifty because it
had been noticed in pilot work that the monotony of the
task negatively influenced the motivation of naive gs.

On the other hand, it is conceivable that the mechanism
underlying the tendency to total assimilation cannot be
that easily modified. Thus, with motivation kept constant
by other means (e.g., payment or change in the task), a
larger amount of practice, perhaps protracted over a number
of days, might have been more successful.

Furthermore, the type of "participation" chosen in
this study may not have been realistic enough to keep §
actively involved in practice. Most §s mastered the task
as required by the instructions after a few trials. For
instance, one § mentioned that the number of trials was
excessive compared with the ease of the task; this cir—
cumstance created suspicion about the real purpose of the
experiment.

It could in fact be maintainei that §s did not acquire
any new responses during the practice phase. Unilateral
apparent movement, for instance, could be conceived of as
a Gestalt in itself which need not be "learned." Repeated
presentation of that Gestalt could then be designated as
"familiarization," but might not transfer to another Gestalt
(the Von Schiller pattern). The motor response made during
practice can be conceived of as a skill already present in

Ԥ}s repertoire or at least capable of being rapidly mastered.

74

This motor response was aroused by instructions at the

same time g was being familiarized with the apparent move—
ment pattern. The saying of the directional phrase too was
brought about by E's instructions. While in the traditional
definition of paired associate learning, § originally

makes errors which are gradually eliminated until a cri—
terion of mastery is reached, the practice period of the 'h
present study hardly contained any of the ingredients
necessary for that definition of learning. The test was

not a discrimination task in the usual sense because there

 

were no physical differences to discriminate. Also, over
a period of time, no forgetting in the usual sense could
be expected to occur.

All these characteristics of practice in the present
experiment may have weakened its relevance for the perform-
ance on the posttest. What practice probably did was to
produce a light set or expectancy for directional conno—
tation in general, much as other factors in the experimental
situation produced the bias reversal effect. The short—
livedness of the state induced by practice is highlighted
by the posttest check. A brief interview period, the
switching on and off of the light helped to dissipate the
effect of practice rather quickly (although a one trial
check on the posttest is not too reliable).

In this context, it should be further noted that
the variable of handedness may have contributed to the

contamination of the results. That handedness and eyedness

75

may be involved in directional choice was discovered early
by De Silva (1926, p. 498f).

The question of right—handedness and right-
eyedness seemed an interesting cue and two lefthanded
students were induced to observe. Experimentation
with equivocal situations demonstrated a preference
for movement in a clockwise direction. In other
words, the directional compulsoriness of the move—
ment in the case of left—handed.gs was found to be
exactly the reverse of that in right—handedygs, a
discovery which indicates that the directional
compulsoriness of clockwise movement or of counter—
clockwise movement may be due to "handedness" or
"eyedness."

As far as attitude is concerned, left practice for right—
handed.§s could be expected to be more of a challenge than
right practice. There seemed to be, in fact, a difference
in attitude between the right and left practice groups
although this difference was not statistically significant.
The attitudes toward the experiment were more positive and
less ambiguous in the left practice group than in the
right practice group and the control group. The attitudes
in the latter group may have been strongly provoked by

the posttests: No practice or experience with apparent
movement could be brought to bear on this rather frustrating

repetition of physically identical trials; and still the

task was to detect differences!

4.3: ‘Thg.tg§t. This brings up the question of whether
the test should have contained actual differences such
that §s with practice had a more realistic chance to use
their experience in the test situation. Color differences

had been found ineffective in pilot work. Perhaps a re—

*3
;\

inforcement of the practiced directional connotation by
means of enhanced brightness or equal stimulus form would
have been more successful. For example, on the pretest,

§s might not have overcome the tendency to total assimila-
tion, although two of the three stimulus members bl—a—bg,
viz. a—bg, were equal in form, whereas b1 was different.
Practice of the direction a-bZ (right) might have endowed
this direction with a special connotation on the posttest.
Also, since §s expressed directional connotations in terms
of "faster" or "further,” they could have been trained to
discriminate actually between "faster" or "further" apparent
movement. The test could have contained stimulus situations
relevant to that training; transfer from unilateral training
patterns to the Von Schiller bilateral pattern could thus

have been studied.

4.4: The rgle 9; awareness. The design of the present

 

study prevented all §s but one from seeing any clear con—
nection between what they perceived as the first experiment
and the second experiment. This precaution had been nec—
essary to counter, at least in part, the argument that the
verbal reports were made up by §s to either help or defeat
g. Bartley (1958) states that:

One of the greatest desires of the naive
observer is to be "right”: that is, to act in such
ways as not to come into conflict with physical
externality, or with custom ... the attempt to be
"right" is injected into laboratory behavior and

often runs counter to laboratory instructions and
invalidates the experimentation. (p. 29)

77

Had the naive §s understood that they were tested on whether
they were able to transfer directional connotations from
practice to posttest, they might have felt tempted to

report not what they saw but what they thought they should
see. Through the rigging of the experiment, an attempt

was made to guarantee that the verbal reports reflected,

in fact, perceptual experience. This attempt seems to

have succeeded.

4.5: .E222231.2£ experimental design. A further discussion
of the present study, given the nature of its results,
should center also around the question of how the experi—
mental design can be improved and what can be learned for
further research. One decisive drawback of the present
design has been the uneconomical way of obtaining unbiased
§s. Instructing §s to see apparent movement was a legiti-
mate way of augmenting the number of usable §s. This pro-
cedure is consistent with what we know about the influence
of set and instructions on perception of apparent movement,
especially the work of Kelly (1935). The percentages of
110 §s reporting apparent movement duriig the screening
phase (82%) and in the posttest interview (88%) do not
differ significantly from those reported by Kelly. He
used flashlight bulb patterns with 60 msec intervals. His
.§s were four hundred students from elementary psychology

classes. About 50% of the gs reported movement spontane-

78

ously when the lights were presented flashing alternately.
When §s were told what to look for, 94% reported apparent
movement. The somewhat lower percentage obtained in the
present study may be due to the somewhat more difficult
task of seeing movement in the Von Schiller bilateral
pattern.

A serious loss of SS (roughly 21%) was brought about
by the Ss' failure to follow Von Schiller's "tendency to
total assimilation" when they reported seeing apparent
movement in the screening phase. Von Schiller had obtained
his optimal effect with flash durations of 500 to 800 msec
and intervals of lO_25 msec. He recommends, however, that
if Ss are not able to see apparent movement under these
conditions, the flash duration and interval be shortened
to 100—200 msec and lO—15 msec respectively. His recom-
mendations were followed after pilot work had indicated
that shorter flash diration and interval tended to maximize
the number of SS seeing apparent movement. But there is
a possibility that these short durations also increase
deviation from the "tendency to total assimilation." The
introduction of a fixation point might possibly reduce the
number of deviations, although it is known to reduce apparent
movement experience.

The most serious loss of §S is, of course, due to
bias in directional connotation exhibited on the pretest.
Only 37% of the total number of SS who saw apparent movement

Were also relatively unbiased and hence eligible for

79

directional practice. This finding is not too surprising

if we recall Willey's (1936) study on individual differences
in directional variations of apparent movement. Both Von
Schiller (1933) and Willey (1936) report that it is possible
to "retrain" biased.Ss. Voluntary control of directionality
in a choice situation is possible, especially if this control
is exerted in terms of motor imagery (Von Schiller, 1933,

p. 18). The difference between Von Schiller's and Willey's
situation and the present study is, of course, that in the
latter case naive observers were used. Nevertheless, there
is significant evidence from the present study that bias
decreases from pretest to posttest whether §s practiced one
or the other direction or not.

The design of the present experiment would have been
generally more efficient had it been an analysis of variance.
Interaction of practice direction, weak and strong bias,
and other variables could have been analyzed in a more

convenient fashion.

4.6: Qefinition g: Conngpatign. In keeping with the

 

preliminary character of this study, response categories
were not prescribed for §. This caution, however, was not
necessary. The majority of SS given the instruction to
make qualitative discriminations tended to adopt spontane—
ously the categories "faster" and/or "farther." These
categories are reported in the literature. Thus De Silva

(1928) lists the following connotations in apparent movement:

80
l. Existentialtiy (continuity of passage, most
real, least inferred)
2. Swiftness (apparent velocity)
3. Smoothness (or jerkiness)

4° Filling of Zwischenraum (greyness, color,
brightness, and7or saturation of Zwischenraum)

5. Liveness (Zwischenraum flashy or glittery)

6. Homogeneousness of ggischenraum (uniformly
lighted or patchy)

7. Outline of Ewischenraum (straight or constricted)

8. Reverse movement (stimulus flashed second appears
to jump back toward the one flashed first)

Other authors, such as Jones and Bruner (1954),
and Toch and Ittelson (1957), found that §S used only
"farther" and "faster" when they reported on qualitative
differences. Haggard and Rose (1944) used S's estimate of
movement magnitude (in inches) as a parameter of their
dependent variable in an experiment on the conditioning of
the autokinetic phenomenon. Von Schiller (l933) divided
qualitative responses into two categories, which he called
"Fassung g" and "Fassung 1," respectively (version g and
version 1). He characterized as version i such responses
as "Sch8ne glatte Bewegung" (beautiful, smooth movement),
"Gute ruhige Drehung" (good, calm turning around), "Klappt
schbn, sicher" (works well, secure), "Einwandfreie, vollkommene
ABewegung" (impeccable, perfect movement). Version y is
exemplified by the following responses: "Buckartiges,
“unnatfirliches Springen" (jerky, unnatural jumping), "Auf_

fkillend schlecht, unerklarlich" (remarkably bad, inexplicable),

8l

"Widerliche, sonderbare Bewegung" (repugnant, strange move—
ment), "Etwas ist nicht in Ordnung" (something is not in
order), "Es geht los, dann ist es platzlich ganz anderswo
and ganz anders ohne fibergang" (it starts and then it is
suddenly somewhere else and something else without tran-
sition).

In comparing the findings of all these authors,
responses like "faster," "farther," "smoother," etc.,
emerge as the more frequent ones. Other responses, if
they referred to movement in a horizontal direction and
not to the stimulus lights, were scored according to Von
Schiller's technique of classifying responses into positive

and negative versions.

5. Summary

Apparent movement has been a focus of interest in
perceptual theory since the early days of the Gestalt
school. Wertheimer's short circuit theory of apparent
movement was early criticized from many quarters. The
discovery of such factors as attitude and set made premature
physiological explanations doubtful and directed attention
toward other determinants of apparent movement. The vari—
ability of experiences under standard external conditions
suggested the necessity of investigating the past history of
an organism. Past experiences, it was postulated, must
interact with the physiological mechanism of movement
perception. Heavier and lighter weighted past experiences
could be eXperimentally manipulated by using meaningful
stimulus material in synchronism with or antagonism toward
deep—seated movement "assumptions." A symmetrical stimulus
pattern (bl-irbz)’ used first by Von Schiller to demonstrate
what he called "tendency to total assimilation," was now
used with meaningful figures to demonstrate the effect of
meaning in offsetting total assimilation.

The next step in this development was to bring the

Concept of "meaning connotation" under laboratory control.

82

83

An attempt in this direction was made in the present
study. The Von Schiller bl‘arbZ sequence was presented to
SS in a pretest with relatively meaningless circular lights.
The pretest was used to single out §s who were relatively
unbiased directionally and exhibited a tendency to total
assimilation. These Ss were given intensive experience
with only a part of the bl—g-b2 pattern. One group looked
repeatedly at the g_b2 sequence to the right, another group
the a—bl sequence to the left. While looking at apparent
movement to the right, or to the left, §s manipulated a
lever apparatus in the corresponding direction. §s also
verbalized the direction of the apparent movement. The
hypothesis was: Unbiased Ss who, in a previous performance,
looked repeatedly at unilateral apparent movement while
actively reinforcing that direction with motor gestures
and verbalization, will on a later performance, exposed to
a bilateral apparent movement pattern, increase the number
of reports with connotation in the direction of practice.
This hypothesis was not substantiated by the experiment.

No specific directional effects were found, although §s
tended to report more connotations in the direction opposite
to that of practice. Interaction of bias reversal and
practice, insufficient intensity of practice and attitudinal
differences between the experimental groups were counted
Eflmong the reasons for the inconclusiveness of the results.
It was, however, found that practice did increase signifi-

Carrtly the overall number of reports with directional

84

connotation in unbiased experimental SS. This was inter—
preted as a heightening of directional expectancy due to
exposure to directionality in practice. More work in the

framework of the hypothesis proposed by the study is neces-

sary.

10.

ll.

12.

Bibliography

Allport, F. H. Theories pf pergeption and the concept
‘9: structure. New York: Wiley, 1955.

Bartley, S. H. Vision: A_study pf its basis. New York:
Van Nostrand, 1941.

Bartley, S. H. Principles pf perception. New York:
Harper & Brothers, 1958.

Blug, A. Neue Untersuchungen uber Scheinbewegungen
bei tachistoskopischen Beobachtungen, g. Psychol.,

1932, '__Z, 290— 324.

Boring, E. G. Perceived movement. In: Sensaplpn and

perception in the history of experimental psgchology.
New York. Appleton-Century-Crofts, 1942, 5 - 02.

 

Brenner, May W. Continuous stimulation and apparent
movement. Amer. g. Psychol., 1953, pp, 494—495.

Brenner, May W. The effect of meaning on the perception
of apparent movement. J. nat. Inst. personn. Res.,
Johannesburg, 1956, p, 125—131.

 

Brenner, May W. The effects of brain damage on apparent
movement. J, Pers., 1956, 25, 202-212.

Brenner, Nay W. The developmental study of apparent
movement. Quart. g. exo. Psychol., 1957, 2, l69—l74.

Brenner, May W. The differential effects on apparent
movement of satiation, brain damage and schizophrenia.
J. nat. Inst. personn. Res., Johannesburg, 1957,
Z, 66.

Brown, J. Directional information and apparent movement.
A note on the paper be Malcolm A. Jeeves and Jerome

A. Bruner. Quart. J. xp. Psychol., 1957, 9, 111-112.

Deatherage, B° H. & Bitterman, M. E. The effect of
satiation on stroboscopic movement. Amer. J. Psychol.,
1952, 6 , 108-109.

85

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

86

De Silva, H. R. An experimental investigation of the
determinants of apparent visual movement. Amer.

g. Psychol” 1926, 12, 469-500.

De Silva, H. R. Kinematographic movement of parallel
lines. J. gen. Psvchol., 1928, J, 550—557.

Dimmick, F. L. An experimental study of visual move-
ment and the phi phenomenon, Amer. J. Psychol.,

1920, 11.. 317—332.

Dimmick, F. L., & Sanders, R. W. Some conditions of
the perception of visible movement. Amer. J. Psychol.,
1929,.4J, 607—616.

Gantenbein, Maria-Martha. Recherche sur le development
de la perception du mouvement avec l'age. Arch.
Psychol., Geneva, 1952, 3}, 197—294.

Gengerelli, J. A. Apparent movement in relation to
homonymous and heteronymous stimulation of the
cerebral hemispheres. J. xp. Psychol., 1948, 38,
592-599.

Haggard, E. A. Experimental studies in affective
processes: I. Some effects of cognitive structure
and active participation on certain autonomic re-
actions during and following experimentally induced

stress. J. egp. Psychol., 1943,‘33, 257-284.

Haggard, E. A., & Rose, G. J. Some effects of mental
set and active participation in the conditioning ofi
the autokinetic phenomenon. J. xp. Psychol., 1944,

fl, 45-590

Haggard, E. A., & Babin, Rachel. On the problem of
'reinforcement' in conditioning the autokinetic
phenomenon. ‘J. exo. Psychol., 1948, 22, 511—525.

Hall, K. R. L., : Earle, A. E. & Crookes, J. G. A
pendulum phenomenon in the visual perception of
apparent movement. Quart. J. exp. Psycho;., 1952,
4, 109—120.

Hastorf, A. H. The influence of suggestion on the
relationship between stimulus size and perceived
distance. J. Psvchol,, 1950, 29, 195—217.

Higginson, G. Apparent visual movement and the Gestalt.
J. exp. ngchol., 1926, 2, 328—352.

Hovland, C. I. Apparent movement. Psychol. Bull.,

1935, 32, 755—778.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

87

Jeeves, M. A., & Bruner, J. S. Directional information
and apparent movement. Qpart. J. exp. Psychol.,
1956, 8, 107—113.

Jones, E. E., & Bruner, J. S. Expectancy in apparent
visual movement. Brit. J. Psychol., 1954, 4 , 157-
165.

Kelly, E. L. The effect of previous experience on the
perception of apparent movement. Psychol. Bull.,

1935, 32; 569-570-

Kilpatrick, J. P. Two processes in perceptual learning.
J. exp. Psyphol., 1954, 4 , 362—370.

Krampen, M., & Toch, H. H. The determination of
perceived movement direction. J. Psychol., 1960,
5Q, 271—278.

Krolik, W. fiber Erfahrungswirkungen beim Bewegungssehen.
Psychol. Forsch., 1935, 20, 47-101.

Levit, M. Some parameters of visual apparent abient
movement. unpublished doctoral dissertation, Michigan
State University, 1957.

Meili, R., & Tobler, E. Les mouvements stroboscopiques
chez les enfants. ,Aqgh. Psychol., Geneva, 1931,
.23, 131—156.

Morris, C. W. Foundations of the theory of signs.
Int. Engycl. unif. Sci., 1938, l, 63-75.

Morris, C. W. Signs languagg and behavior. New York:
Prentice—Hall, 1946.

Neff, W. S. A critical investigation of the visual
apprehension of movement. Amer. J. Psychol., 1936,
48, 1-42.

Neuhaus, W. Experimentelle Untersuchung der Schein—
bewegung. Arch. geg. Psychol., 1930, 25, 315—358.

Ogden, C. K., & Richards. T. A. The meaning pf meaning.
New York: Harcourt, Brace & Co., 1946.

Orlansky, J. The effect of similarity and difference
in form on apparent visual movement. Arch. psychol.,

1940, 35, 1-85.

Osgood, C. E. Method and theory J3 experimental
psychology. New York: Oxford University Press, 1958.

 

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

88

Osgood, C. E. The nature and measurement of meaning.
Psychol. Bull., 1952, 49, 197-237.

Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. The
measurpment pf meaning. Urbana: University of
Illinois Press, 1957.

 

Pieron, H. Remarques sur la perception du mouvement
apparent. Annfie_psychol., 34, 244-248, 1933.

Shapiro, M. B. A preliminary investigation of the
effects of continuous stimulation on the perception
of "apparent motion." .Bglp. J. Psychol., 1954,

12, 58—67.

Smith, K. R. Visual apparent movement in the absence
of neural interaction. Amer. J. Psychol., 1948,
61, 73-78. ‘

Smith, K. U. The neural centers concerned in the
mediation of apparent movement vision. ‘J. _ex .
Psychol., 1940, pp, 443-466.

Smith, K. V. & Kappauf, W. E. A neurological study
of apparent movement vision in the cat. ‘J. Gen.

Psychol., 1940, 23, 315-327.

Steinig, K. Zur Frage der Wahrnehmung von Zwischen-
stadien bei stroboskopisch dargebotenen Bewegungen.

Z. Psychol., 1929,'199, 291-336.

Teuber, H. L., & Bender, M. B. Alterations in pattern
vision following trauma of occipital lobes in man.
J. Gen. Ppychol., 1949, 40, 37-57.

Toch, H. H., & Ittelson, W. The role of past experience
in apparent movement: A revaluation. Brit. J.

Psychol., 1957, 47, 195-207.

Von Schiller, P. Stroboskopische Alternativversuche.
Psychol. Forsch., 1933, J1, 179-214.

Wapner, S., & Werner, H. Ppppeptual develgpmegp Ea
invegtigation pithin the framework p£_sensory-tonic
field theory. Worcester, Mass.: Clark University
Press, 1957, 27-36.

 

 

Wapner, 5., Werner, H., & Chancler, K. A. Experiments
on sensory-tonic field theory of perception: 1.
Effect of extraneous stimulation on the visual
perception of verticality. J. Egp. Psychol., 42,
341-345, 1951.

54.

55.

56.

57.

58.

89

Weiner, M. Perceptual development in a distorted room:
A phenomenological study. Psychol. Monogr., 1956,
19 (16) (Nr. 423) 38 p.

Werner, H. Motion and motion perception: A study of
vicarious functioning. J. Psychol., 1945, J9, 317-
327.

Werner, H., & Thuma, B. D. A deficiency in the per-
ception of apparent motion in children with brain

injury. Amer. J. Esychol., 1942, 55, 58-67.

Wertheimer, M. Experimentelle Studien fiber das Sehen
von Bewegung. ,J. Psychol., 1912, 6J, 161-265.

Willey, C. J. Directional variations of apparent
movement. IJ. Exp. Psychol., 1936, J9, 544-673.

APPENDIX I

Epecific Procedure and Instrgctions.

(a) Preliminary instructions. As‘E entered the
room, E introduced himself and invited E to sit down in
the experimental chair. E was told to keep his head in
the headrest, to watch the center of the screen, and to
find a comfortable position. Then, pointing at the levers,
E said:

You will have to use these levers during one

of the experiments. Put the apparatus in such a

position that you can comfortably move the levers

out to the side until they hit the switch. Try to
do this without moving your head in the headrest.
E waited until E had found a comfortable position and

then continued:

I will now turn off the light and give you
some time to get used to the dark.

E turned off the light and said:

I will use this time in which your eyes get
used to the dark to prepare the experiment. But
first, would you please spell your name for my
records.

E wrote down the name. Then he asked:
Are you comfortable?
After a while, E continued the instruction:
(b) mama... g. the «......»1...:_._-;«
Now, let me tell you a little about these

experiments. You will actually participate in two
seperate studies. The first one is designed to

90

91

measure your ability to coordinate your perception
with a hand movement and with your verbal reaction
(to perceive time intervals correctly). The second
is designed to test how accurate you are in per-
ceiving small difference in quality. In both (one)
experiments lights will be flashed on the screen in
front of you. A set of electronic timers regulates,
to very small fractions of a second, how long the
lights are on the screen. With this equipment, we
can produce apparent movement. Do you know what
apparent movement is?

If E indicated that he knew what apparent movement was,
‘E pressed for an example by saying:

Yes; can you give me an everyday practical
example?

If E's example did not demonstrate knowledge of the
phenomenon or if he had responded negatively to E's question,
it was explained to him as follows:

Well, apparent movement is a technical term
in experimental psychology. And I would like to
give you a practical example of what it stands for.
Have you ever seen those neon arrows flashing back
and forth at a filling station or in outdoor adver-
tising?

E waited for an answer, then continued:

Now, when you looked at these arrows, did
you have the feeling that they made some kind of a
movement as if they were pointing at something or
flashing back and forth?

.E waited for an answer, then continued:

Well, this particular impression of movement
you had is what we call apparent movement. I guess
you see now that apparent movement is a similar,
but p93 the same kind of experience which you would
have when you look at a ping pong ball going back
and forth. Only the neon arrows at the filling
station appear to be moving. That's why their
movement is called 'apparent.‘ But, let me show
you an example of what I mean. Would you please
watch the screen and just tell me what you see.

92

E showed the three test slides. First test slide: apparent
movement left to right. E asked:
What did you see?
‘E waited for an answer. If E reported apparent movement,
E showed the second test slide: Apparent movement right to
left. He said:
What about this one?
E waited for an answer. If E reported apparent movement,
E continued:
And this one?
E showed the third slide: Apparent movement from center
to both sides. After E had reported apparent movement, E
continued instructions. If E did not report apparent
movement consistently, E asked:
When you looked at those neon signs, did
you really have the impression that something was
moving?
If the answer was negative, E marked the record with a
minus Sign. If the answer was positive, E continued:
Then look at this again and tell me whether
what you see is similar to that experience with
the neon signs.
‘E repeated the last test slide. If the answer was again
negative, E marked the record with a minus sign. If the
answer was positive, E marked a plus sign. Then E concluded
the introductory instructions for all Es as follows:
I have shown you these examples. Do you

think that we have now a common understanding of
what I mean by apparent movement?

93

E waited for the answer and then continued the instructions:

Then, let me tell you now how we'll proceed.
I think that, before taking the two experiments,
you should really have the opportunity to get familiar
with your task. So we'll start out by rehearsing
the first experiment. Then we'll rehearse the second
experiment. After both rehearsals, you will take
the first one and then the second one. After the
experiments, then I would like to ask you still a
few questions.

(c) "Ephearsal" 2i practice (experimental groups)

Now let's rehearse the first test. As I
said, this test will measure your reaction time,
and your ability to coordinate your responses. So,
we'll first try out what I mean by coordination of
responses. Would you please take the right (left)
lever in your right (left) hand. O.K.? Now release
the lever; let it snap back. O.K.? Now I would
like you to make the same lever movement together
with apparent movement to the right (left). I will
flash light on the screen and you start the lever
movement when you see the apparent movement start.
Try to hit the switch and release the lever as soon
as the apparent movement is over. Any questions?
Ready?

E prepared the apparatus for apparent movement to the

right (left) and gave one or two trials until E understood

the task. Then he continued:

What you have just tried is the coordination
of two responses, watching the apparent movement and
moving the lever. In the test, I will complicate
the task a little more by adding a third response.
This will be a verbal response. Everytime you see
the apparent movement and you move the lever, you
also say the phrase, "To the right (left)." That
is all you have to do in the first experiment. I
will count the number of trials you need to come
up with a good and stable reaction time and coor-
dination. Any question?

For the control group, the instructions were:

Now let's rehearse the first experiment. As
I said, this experiment will measure how accurate
you are in estimating the duration of time intervals.
All you have to do here is to give me your impression

94

how long a certain time interval appears to you.
After I say "ready" you will hear the "click" of a
stopwatch which is a signal for you that the interval
has begun. Then you simply sit back and relax until
you hear the second "click" which marks the end

of that interval. At the end of the interval, I

will say "now", and you give me immediately your
impression how long the interval appeared to you.

The intervals may be longer or shorter. Do not

count or reason your answer out. I am only interested
in your first impression. Do you have any question?
Now let's try it once. Are you ready? (Clicks)

"NOW ! II

'E concluded the first rehearsal for experimental and control
group alike:

That is all you have to do in the first
experiment. Do you have any questions?

After this question ending the first period of
rehearsal, E waited for an answer. If there were any
questions, E answered them by paraphrasing the instructions
or by saying "Yes" or "No." Then E continued:

O.K. Then let's rehearse now the second
experiment.

(d) "Rehearsal" _£ socond_expepiment (pretest)

This is an experiment in perception of small
differences. (So you can for the time being forget
about levers.) I will flash light which should
give you an apparent movement splitting from the
center to the right and the left. I am going to
measure whether you can perceive small differences
in quality between the two halves of the movement.
Let me set the apparatus for you.

.E set the apparatus and then continued:

The apparatus is now set in such a way that
you should be able to see these small differences
which could a pear on each trial. Since these
differences sfiould be there, make no attempt to
guess, but just tell me what kind of differences
you see between the right and the left half of the
movement. Every detail you should observe is very
important. Please report it at once. If you do
not see any difference admit it by saying ”can't

95

tell.” So all you have to do now is to look very
hard comparing the two halves of the movement as to
their quality. And please rely always on your
very first impression. Any question?
If E had questions, E answered them by paraphrasing the
instructions or by saying "Yes" or "No." The instructions
continued:

Since this test is not as easy as the first
one, I would like to give you a series of practice
trials. Are you ready?

E showed ten pretest trials. Each presentation was preceded
by "ready" and a three-second pause after "ready" before

the trial was given. Then E waited for a response. If,
after the first trial, the response showed that E did not
understand the instructions, E paraphrased the relevant

part of the instructions. If only bilateral movement was

reported, E might have probed to obtain more detail by

saying:

Was there any difference between the two
halves of the movement?

After the pretest,, E said:

Well, do you now have a good impression how
the second test will be?

.E waited for an answer, then he continued:

(e) Practice (experimental groups only).

 

Alright, then let's stop rehearsing and
start out with the first test.

.E prepared apparatus for practice and said:

96

Remember, you have to watch apparent movement
to the right (left), say the phrase "to the right
(left)fl and move the lever to the right (left).

It is important that you hit the switch everytime
because the switch is connected to a recorder.
And now I will start counting the trials. Ready?

E presented the first group of twenty-five trials. The
trials were given at regular intervals, After each response,
E waited three seconds before presenting the next trial.
If E did not respond correctly, E let E know that he had
detected the mistake and asked for correction. After the
first group of trials, E said:
Are you getting tired?
E waited for an answer and then continued:
Well, this is a pretty monotonous task.
But I think you could make your verbal response
(your lever movement) still closer together with
the apparent movement. Would you mind making another
set of trials?
E gave the second set of twenty-five trials in the same

fashion as the first and then continued:

That's enough now. You did a very good job.
We have only one more experiment to go.

,For the control group the same instructions as in the
rehearsal were repeated.
(f) Posttest.

Let me repeat the instructions of the second
test. Remember, this is an experiment in perception
of small differences (so you do not move the levers).
I will flash light which should give you an apparent
movement splitting from the center to the right and
to the left. I am going to measure whether you
can perceive small differences in quality between

the two halves of the apparent movement. Now, let
me set the apparatus for you.

.5 set the apparatus and then continued:

97

This apparatus is now set in such a way that
you should be able to observe these small differences
on each trial. Since the differences should be
there, make no attempt to guess, but just tell me
what kind of difference you observe between the
right and the left half of the apparatus movement.
Every single detail you observe is important. If
you do not see any difference, admit it by saying
"can't tell." So all you have to do now is to look
very hard comparing the two halves of the apparent
movement on their quality. And please rely always
on your very first impression. Ready?

E showed ten posttest trials. The trials followed three
seconds after E had announced them by saying "ready."
After each trial, E waited for the. response. IflE still
showed misunderstanding of the instructions, E might here
have paraphrased the relevant part of the instructions in
between the trials. If the movement was reported without
difference, E might probe deeper by asking:

Was there any difference between the two
component movements?

(g) Posttest intepview and check. After the last

 

posttest trial, E turned on the light and said:
Well, that's all. Thank you again for your
cooperation. By the way, did you find this last
test difficult?

E waited for an answer and then continued:

That is interesting! Do you mind if I ask
you a few questions?

‘E_waited for an answer and then continued:
When you saw these differences between the
two movements, which you reported as smoother (faster,
brighter, etc.), how did this really look to you?
Can you give me some more details of what you saw?

.E recorded the answer and then continued:

98

You know, I was thinking of what you said
before about the movements. Can I show you one
more trial? Just tell me what you see. O.K.?
E waited for an answer and then turned off the light and,
after a short period of dark-adaptation, said:
Ready?
Then E showed the posttest trial again and waited for a
response. If E did not mention movement, E said:
Did you see movement?

If E did not report differences, E asked:

How about the movement to the right and to
the left? Did you see any difference between them?

E repeated the trial in order to obtain a clear response.
After the response was obtained, E turned on the light
again and said:
Well, it was a pleasure to work with you.
Before you go, would you please write a few comments
about the experiments in this form. Just continue
the phrase which you find begun and write as much
as you want.

E answered questions about the questionnaire if necessary

before dismissing S

APPENDIX I I

CHI SQUARE TABLES

Table 7

Posttest increase of unbiased vs. biased Es

 

— ——

 

 

 

 

 

 

Es increased other I both
1 l
unbiased 23 17 40
biased 5 18 23
totals 28 35 63
chi square, 7.563 1 df. .01) P

99

 

100

Table 8

Posttest increase of unbiased Es reporting directional

 

 

 

 

 

 

 

connotations
Es increased other both
experimental 17 6 I 23
control 6 11 l 17
I -—
totals 23 17 4O
chi square, corr., 4.490 1 df .O5>P>.02

101

Table 9

Posttest increase of biased Es reporting directional

 

 

 

 

connotations
Es increased other both
experimental 3 12 15
control 3 5 8
__. 1___ _.._ ‘9‘
totals 6 l7 5 23

 

 

 

 

———— _— _.__

chi square, corr., .17 1 df .7>P>.5

102

Table 10

Increase in connotations to the right, Es with practice

to the right vs. Es with practice to the left

— —— ——-———

__ _ _

 

 

“

 

j—fi—_—‘-—_,

chi square, corr., 1.483 1 df

 

 

_ increased rt. other both
practice rt. 2 7 9
practice 1f. 8 6 14
totals 10 13 23

103

Table 11
Posttest increase of unbiased.Es reporting

connotations to the left after right practice

._. —
‘7

 

-
‘— —— —

 

88 increased 1f. ‘ other both

 

 

practice rt. 6 ‘ 3 9
controls 5 12 17

 

totals ’ 10 16 ‘ 26

 

 

 

 

 

chi square, corr., 2.056 1 df .2>P>.1

104

Table 12

Posttest increase of unbiased Es reporting

connotations to the right after left practice

 

 

 

 

 

Es increased rt. other both
practice 1f. 8 6 14
controls 3 14 17
totals 11 2o 31

chi square, corr.,

 

3.648 1 df

 

_——'

.1>P>.05

 

105

Table 13

Interaction of training and bias in slightly

biased.Es

increase in direction

 

 

 

 

Practice I—OI practice oppos1te praCtice both
against bias 5 0 5
in direction of 1 8 9
bias

totals 6 8 l4

 

 

 

 

chi square corr., 7.058 1 df

106

Table 14

Interaction of training and bias in strongly biased Es

——

 

 

 

 

Practice increase in direction

of practice opposite practice * lboth
against bias 5 1 6
in direction of 2 5 7
bias |
totals 7 6 13

 

 

 

chi square, corr., 2.006 1 df

 

 

 

Interaction of training and bias in all biased Es

107

Table 15

 

 

practice

increase in direction

 

practice

opposite practice

both

 

against bias
in direction of

bias

 

—— —

totals

10

13
14

 

 

 

 

27

 

————'~

chi square, corr., 10.858

108

Table 16

Proportion of completely unbiased and slightly

biased.Es in the relatively unbiased

experimental groups

 

 

 

 

 

 

 

 

Es practice practice both
rt. lf.
compl. unbiased l 3 3 6
slightly biased ( 6 10 17
totals 9 14 23
chi. square, corr., .022 1 .90 >P>.80

109

Table 17

Proportion of completely unbiased and slightly biased Es in

the relatively unbiased experimental and control groups

 

 

 

 

 

 

 

 

Es practice control both
'86mpletely unbiased 6‘—— w 7—5 Ill
slightly biased 17 12 29
E.,... — I 23 _ 17 "J."
8111 squar—e, c‘orr.,II016 _ 1 df .95>P> .90

110

Table 18

Difference in attitude between experimental and control

 

 

 

 

 

 

groups
attitudes practice no practice totals
. .1.

positive 14 6 20
other 2 9 11

I
both l l6 15 31

I ll

 

 

chi square, corr., 5.696 1 df .02>P> .01

111

Table 19

Difference in attitude between experimental groups

 

 

 

attitudes _ practice
‘3———r
‘ right left both
|
positive 3 l 11 * 14
other 1 1 2
totals ‘ 4 12 16

 

 

 

 

112

Table 20

Difference in attitude between.Es practicing to the left

and.Es without practice

 

~—

 

 

 

 

 

 

 

attitude practice If. no practice both
p8sitive b I_Il_—‘——I—‘— I—I6 17
other g l 9 10
tbtalsfi_ ‘ _—N' 12 I—— 15 ‘— 27
chI—square, corrI, 5.577 Idef .02;P;:.01 ‘—

 

Posttest check

113

Table 21

 

 

 

 

 

 

 

 

 

Es directional conn. no conn. both

experimental 12 ll 23

l

1
controls I 3 14 17
totals 15 25 40
__, __;:_ ______ ___ __‘_ l‘_fi'_;_
chi square, corr., 3.607 1 df .10>P>.05

   

uriiiyyyiyimy