”I

l
1

MW“(“WNW")

HWWIWI

144

 

2:2: 2.222.222 22 32:22... '
;.<<<;<22<22z;m< <22 CE: A2252 5 max: :22
2:222, A2 22c: A232
LAEELQRC 2232022522

72222233 {av {he Doqm'a <22? E5“? «‘2.
EEC” $23 3,22 22;: " 321““2'2“

LgiJol‘ A52: 2-31! fixad‘uo'ai

Marg are L L. Connolly
1%?

THEE“:

  
 

22“”
LIPR JURY

N‘iChxzw 3 J"

UHiVL‘lmC,’

ABSTRACT

THE EFFECTS OF STIMULUS INCONGRUITY _ u
on CHILDEEI'S‘LooxiNc TIRE, AFFECT, Ans LAbELING
RESPONSES
By

Margaret R. Connolly

This study is an exploratory investigation of perceptual explora-
tion and responsiveness to incongruity by young children. An incongruous
stimulus is defined as a stimulus with components that the observer ’
perceives as incompatible on the basis of past experience.

The g; were 20 kindergarteners, 20 second-graders, and 20 fourth-
graders. Each §_looked at the congruous and incongruous stimuli in a
large picture book. The book contained 15 outline drawings of congru-
ous animals and 15 of incongruous animals. The congruous figures were
familiar animals; the incongruous drawings were figures designed with
the head of a familiar animal and the body of a car, airplane, or a
second familiar animal.

The §.and two observers (gs) recorded onto an event recorder the
length of time'g looked at each picture, the number of times his expres-
sion changed while looking, and the duration of each expression change.
One‘Q’also noted the characteristic expressions of each §, Immediately
following the test session each §_was asked to give each picture a name.

Duration of looking and duration of expression change were signi-
ficantly longer to incongruous drawings than to congruous drawings

(ps‘(.0.01). There were significantly more first-expression-changes

Mhrgaret R. Connolly

to incongruous than to congruous figures (g,<’0.001). The percentages

of kindergartenera, second-graders, and fourth-graders who acknowledged
both components of the drawings were 65, 75, and 95, respectively. No

age effects were significant.

The older children's more frequent acknowledgment of the two incom-
patible components (i.e., incongruity) may be seen as evidence of in-
creasing perceptual flexibility, specifically the ability to 'decenter‘,
with age. However, the present study in-cluded no independent assess-
ment oflgs tendency to center so that this interpretation is merely
suggestive for further research. The expression change results suggest
that affective responses are an effective response measure for preschool
and elementary school children in studies of recognition of stimulus
incongruity. The looking time results tend to support previous findings
that children are more responsive to incongruous stimuli than to familiar

stimuli.

APPrOVQdAh“ M
pm 4:? W I265

 

 

 

THE EFFECTS OF STIMULUS INCONGRUITY
ON CHILDREN'S LOOKING TIME, AFEECT, AND LABELING

RESPONSES

by

Margaret R. Connolly

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

MASTER OF ARTS

Department of Psychology
Summer, 1969

$55357

/0/,22/(o7

To Dick

11

ACKNOWLEDGMENTS

Appreciation is owed to many people who helped in the prepara-
tion of this thesis. I wish to thank Dr. Lauren Harris, adviser and
committee chairman, for his continuous guidance and Drs. Ellen Strommen
and John Paul flexinney for their criticism and assistance. I am grate-
ful for the assistance of Bernard White who served as an observer, and
tabulated the results, and Merrill MUtler who designed the computer
program for the statistical analysis.

I would also like to thank the administration and students of the
Spartan Village School and the families of Spartan Village and Cherry
Lane, Michigan State University, for their willing participation in

this investigation and a pilot study.

111

TABLE OF CONTENTS

List of Tables

References

Appendices

iv

vi

vii

25
32
46
55
59‘

Table

Table

Table

Table

Table

1.

2.

3.

5.

LIST OF TABLES

Page
Summary of Analysis of Variance of Duration
of Looking Time to Incongruous and Congruous
Allin]. Drainss O O O O O O O O O O O O O O O O O O 33
Mean Duration of Looking Time to Incongruous and
Congruous Figures as a Function of Grade in School
‘ndrriaIBIOCR.seeoeeeeeeeeeeeee 34
Frequency Distribution of the 8 Possible Patterns
of Response for each Grade Level. . . . . . . . . . 37
Percent of Total Looking Time to Incongruous Figures
as a Function of Trial Block and Grade in School. . 38
Summary of Analysis of Variance of Mean Duration
of Expression Change to Incongruous and Congruous
Aniulnrflin8.aeeeeeeeeeeeeeeeee 40

Figure 1.

Figure 2.

Figure 3.

Figure 4.

LIST OF FIGURES

Page
Number of‘gs (per trial block) with longer looking

times to incongruous than to congruous figures. . . 36

Mean duration of expression change as a function

0f book 0 O O O O O O O O O O O O O O O O O O O O O 41

Number of subjects whose incidence of first-expression-
change to incongruous figures equaled or exceeded the

incidence of expression change to congruous pictures 43

Number of children in each grade acknowledging one.

or both components of the incongruous drawings. . . 45

vi

LIST OF APPENDICES

Appendix A. Complete Set of Pictures Used in the Study.

Appendix 3. Order of Pictures Among the Four Books.

Appendix C. Individual Subject's Preference (Longer Mean
Duration of Looking Time) for Incongruous than

for Congruous Figures by Trial Block.

vii

INTRODUCTION

One of the most characteristic features of living creatures in
the waking stateis their selective attentiveness to external stimuli.
They attend to some features of their environment; they ignore other
features. Since the early 1900's, psychologists have shown increasing
interest in selective attention and particularly in the question, what
are the determinants and correlates of stimulus selection? Several
factors have been identified as effective elicitors of attention. In
general, they can be classified either as physiological variables or
as environmental variables.

Physiological variables are variables associated with the arousal,
or drive state, of the organism. These variables have been investigated
quite diligently by psychologists as well as by biologists. For example,
the selection of food by a hungry animal is seen as the selection of a
stimulus that will satisfy nourishment requirements; similarly, the se-
lection of a quiet, dark room by a tired man is the selection of a
stimulus situation that will provide him with a place to rest. Because
hunger, rest, thirst, respiration, and elimination are basic physical
needs, the explanation of stimulus selection in terms of physiological
variables usually stresses the biological significance, or life-preserving
function, of the selection. The interesting and from the biological stand-
point, somewhat pussling fact, however, is that animals continue to be
active even after fulfilling basic needs. Activity in the apparent
absence of the operation of life-sustaining drives is often called
’exploratory' or 'curiosity' behavior.

Various factors have been found to influence curiosity. The current
study is concerned with one such factor, 'incongruity'. Incongruity is

l

2

the perceived incompatibility between or among stimulus components. It
has received little attention in the research on curiosity. Typically
'investigation' and 'manipulation' are employed as the indexes of
curiosity; however, observation indicates that affective and verbal re-
sponses may be reliable measures of the perception of incongruity. No
research has attempted to compare the effectiveness of an investigatory
response such as looking time with affective and verbal responses. The
current study is such an attempt.

Before further discussion of incongruity and the response measures,
some general considerations of the history of research on curiosity be-

haviors and factors influencing curiosity behavior will be presented.

Theoretical Backggound

Since the early 1900's there have been several explanations of
curiosity, or exploratory behavior, all of which seem to stem from two
theoretical positions. 0n the one hand, curiosity has been explained as
learned, or conditioned, response no different from other such responses.
0n the other hand, curiosity has been described as a basic drive or moti-
vational state. Most psychologists who have studied curiosity have come
to refer to it in the latter sense for at least the two following reasons:
(1) Because it is among the earliest of the infants responses, it seems
unlikely that it could be learned. (2) Curiosity behavior does not seem
to conform to the usual conditioning principles; for example, the strength
of the response elicited by a particular stimulus steadily weakens as the
incidence of the stimulus presentation increases so that the response

never recovers its initial strength.

Berlyne.

Berlyne (1960), long in the vanguard of research on curiosity and
exploratory behavior, argues for a drive reduction explanation of
curiosity. In so doing he asserts that drive reduction is positively
reinforcing. However, he does not maintain that drive arousal is a
negative condition; indeed, it too may be a rewarding state.

Berlyne has tried to identify and relate the presumably environmen-
tal variables with drive arousal and the behaviors associated with drive
reduction. He labels "collative properties" those seemingly environmen-
tal variables that affect the observer's attention to and selection of
stimuli, while "investigation" and "manipulation" are the specific
curiosity or exploratory behaviors that effect arousal reduction.

The collativa properties (so called because their effects seem to
be additive or multiplicative) are inovelty', 'complexity', 'uncertainty',
'change', 'surprisingness', 'conflict', and.'incongruity' (p. 44). H
Berlyne has attempted to differentiate among these variables through the
following characterisations:

(1.) Surprisingness is the incompatibility between the stimulus

and the subjects' expectation, where expectation refers to the act or
state of waiting for or looking forward to an event that is about to
happen.

(2.) 'Incongruity is the arousal of an expectation that is disappointed
by simultaneously perceived elements or by elements perceived in close
succession.

(3.) Change is a function of the degree of dissimilarity between
successive elements in a series and the speed with which dissimilarity
occurs.

(4.) Novelty is a combination of incongruity, change, and surprisingness.

4

It is an inverse function of frequency, duration, and recency of pre-
sentation of stimuli. Novelty, as well as its components, is defined in
terms of the organism’s perceptual history.

(5.) Uncertainty is a function of incompatible, simultaneously aroused
expectations.

(6.) Conflict occurs with the simultaneous arousal of two or more com-
patible responses.

(7.) Complexity is a function of intra-stimulus variety. It increases
with increasing numbers of distinguishable, dissimilar elements. Com-
plexity is relatively independent of the subject's past experience and
therefore can be determined objectively by quantifying the elements of
the stimulus.

Berlyne (1960) has not identified any unique behaviors that are
called 'curiosity' or 'exploratory' responses. He has characterized
curiosity by reference to the following responses: 'orientation', e.g.,
toward the stimulus; 'approach', e.g., walking toward the stimulus;
'investigation', e.g., looking longer at the stimulus; and ’manipulation',
e.g., touching, holding, turning the stimulus. Orientation and approach
responses seem to be associated with an reinforced by drive arousal.
0n the other hand, investigation and manipulation, the more specific
exploratory behaviors, are associated with and reinforced by arousal re-
duction.

Berlyne's theory predicts that some combination of the collative
properties produces maximum stimulus attractiveness and that arousal is
associated with the perception of those specific combinations. The ob-
server’s total level of arousal is gradually increased and he approaches
the arousing stimulus; he explores the stimulus until the level of

arousal returns to its optimal level. The organism, then, is reinforced

5

as a result of arousal reduction. Some arousing stimuli may be avoided.
Other stimuli go unnoticed because the arousal level remains below the
response-activating threshold. Berlyne predicts a quadratic relation
between strength of the collative properties and stimulus attractiveness.

To support the contention that curiosity is a drive that is positively
reinforced by the performance of certain activities that promote drive re-
duction, Berlyne appeals to the similarity of the stimulus-response re-
lationships of curiosity and those of the life-sustaining drives such
as hunger. However, exploratory behavior presumably differs from homeo-
static behaviors insofar as the stimulus that initially elicits a
curiosity response will not elicit that same response on all future
occasions, including situations in which presentation follows a period
of stimulus deprivation. Berlyne seems to suggest that the curiosity
behaviors are.associated with the stimulus dimension, e.g., 'novelty',
rather than with the stimulus object and that the effect of repeated
exposure to 'novelty' parallels that of the life-sustaining stimuli.
Organisms do learn to respond discriminatively to such stimulus character-
istics as color and sise independent of the object. Likewise, they may
learn to respond discriminatively to the collative properties as dimen-
sions independent of objects.

Berlyne has proposed two types of curiosity: "perceptual" and
"epistemic". Perceptual curiosity is a state of high arousal that is
generated by insufficient stimulus information to the sensory systems
and is relieved by "specific exploration" (p. 195). For instance, a
picture of an animal with an elephant's head a a turtle's body (i.e., an
incongruous animal) might be looked at longer than a picture of an ele-
phant. Perceptual curiosity strongly implies overt behavior. Epistemic

curiosity, on the other hand, is a state of arousal engendered by

6
conceptual conflict or what Berlyne calls the "incompatibilities be-
tween symbolic responses" (p. 283). It implies covert behavior, and
as such probably would have to be mediated by language. The present
study deals with perceptual curiosity and incongruity.

Berlyne proposes that the organism approaches source of arousal
induces the organism's approach when the combination of collative prop-
erties is most attractive or when it induces conflict. However, he has
stated neither the critical combinations of collative properties nor the
necessary and sufficient quantity of each property that must be present
to elicit approach. Also, Berlyne has not defined curiosity independent
of the situation in which it occurs. Cantor (1963, p. 3), has criticized
Berlyne on the former point, i.e., for maintaining that conflict (engen-
dered by these stimulus dimensions) elicits approach and exploration
rather than avoidance. Brown (1961) has criticized Berlyne for not
defining exploratory behavior independent of its occurrence in a particu-
lar situation. Both Brown and Cantor suggest that Berlyne's theory is
circular because the collative properties lack independent operational

definitions.

fiebb,

Hebb's associationist theory (1949) also speaks to the issue of
curiosity and stimulus selection but seems to emphasise learning
("expectancy" and "prediction") rather than drive reduction or arousal.
Hebb is concerned with the neurological concomitants of learning, e.g.,
the hypothesised construction of "cell assemblies" and "phase sequences".

According to Hebb (1949), cell assemblies are systems of functionally
"associated" neurons. Neurons that are active simultaneously or in

immediate succession "tend to become 'associated', so that activity in

7
one facilitates activity in the other" (p. 70). A phase sequence is
a cell assembly or a series of assemblies with all of the intervening
motor elements.

With respect to perception Hebb's theory predicts the disruption
of previously established cell assemblies by the unexpected. For example,
a particular triangle 't' has angles a, b, and c. Each time the subject
looks at the angles in a specific sequence, say a-b-c, the cortical
associations among the three angles are strengthened in precisely that
sequence such that when the triangle 't' is perceived again it is likely
to be scanned the same way, a-b-c. If, in fact, an angle 'a' which has
been associated only with triangle 't' is common also to triangle 'x',
the observer, on seeing x', may expect and anticipatorily scan for b and
c. Angles b and c do not follow, the subject's expectation is defeated,
and he sees some other angles instead. According to Hebb, a "defeated
expectation" produces a neurological disruption, one of the consequences
of which is exploratory or curiosity behavior.

The collative properties may be associated with defeated expectations
in still another way: a particular person 'p' has special characteristics,
e.g., blonde hair, glasses, and a white laboratory coat with a red collar.
Each time the subject looks at the person he sees these specific features,
and the cortical associations among the characteristics are strengthened.
If the laboratory coat with the red collar, previously associated only
with person 'p', is now worn by visitor 'v', the subject, on seeing the
laboratory cost, may expect and anticipatorily look for the blonde hair
and glasses. The visitor has blonde hair but no glasses, so the subject
does not see what he expects on the basis of his past experience. Again
the discrepancy between the perception and the expectation would lead to

the disruption of cell assemblies and thence to exploratory behavior.

8

A very large discrepancy might lead to disruption of very intense asso-
ciations among cell assemblies, and consequently to either fear or avoid-
ance behavior.

The associationist theory explains curiosity by referring to hypothe-
tical neurological activities. Though there seems to be some evidence
for neurological activity (e.g., changes in CNS activity as indicated
by alpha rhythm seem to be correlated with perception of collative
variables such as novelty) the nature of the neurological changes is
unknown. Until psychologists have the equipment to study the exact

nature of these changes, Hebb's theory must go unsupported.

Piaget.
Piaget (1952) deals with exploratory behavior from a still different

point.of view, one that we will call "cognitive-developments1”. This view
sees exploratory behavior asdependent upon learning but not necessarily
as learned behavior 225,33. Piaget's theory deals with the development
of conceptual schemas by adaptation of new (novel) stimuli. Adaptation
involves the concurrent processes of assimilation and accommodation. fl
Curiosity seems to play a role in the assimilation phase of adaptation.
According to Piaget, assimilation occurs when an organism encounters
an object and responds to that object as he responds to a similar, familiar
object. The behavior may be adapted to the object's structure but in-
appropriate to its function: for example, a child may grasp and try to
bite a ball. The child's behavior, at this stage in adaptation, might
be labeled 'exploratory'. Accommodation is the phase in which the organism
has differentiated the new object from other similar objects and has be-
gun to develop behavior patterns specific to the object. He is inte-

grating the stimulus into his mental network of schemes as he is developing

9
the differential responses. The stimulus eventually becomes familiar
and is a standard, so to speak, with which future stimuli are compared.
It is by reference to familiar stimuli that the future stimuli are per-
ceived as 'novel', 'incongruous’, etc.

Adaptation may be slow so that the stimulus changes before it
has been fully integrated. Modification of a stimulus that has not been
integrated completely into the existing schema would probably prolong
the assimilation response pattern. However, change in an object that
the organimm has integrated into its cognitive schema and for which he
has a specific response repertoire available, may disrupt the specific
response pattern and cause a reversal to the trial-and-error behaviors
characteristic of assimilation. In other words, the stimulus will be
explored. Stimulus change could be induced by almost any of Berlyne's
collative properties.

Piaget's theory deals with schema development and reorganisation
that occur simultaneously with adaptation of a stimulus. As the child
has more experiences and applies the old schemas to new experiences, the
schemas are reorganized and become more complex. During the "perceptual-
motor" stage the child may develop rudimentary schemas for classes of
objects such as 'round', 'new', 'familiar', as well as schemas for the
objects themselvas, e.g., 'face', 'ball'. It seems that the young child
quickly learns to manipulate round objects differently from the way he
handles square objects. This different behavior may suggest the formation
of a schema for 'things-that-are-round'. On the other hand, infants and
young children frequently have difficulty discriminating food objects.
This problem.might arise as a function of the variation in size, shape,
and texture of food objects; that is, the variety of perceptual cues

for the object category 'things-to-eat' precludes easy discrimination

10
of these objects. There is no unique discriminable stimulus dimension
for this important stimulus class.

If early cognitive growth is really the development of stimulus-
dimension schema rather than a stimulus-object schema, it implies that
infants should respond differentially to objects with familiar stimulus
dimensions, since these have discriminable cues, and similarly, should
not differentiate among objects with predominately unfamiliar stimulus
dimensions.

Piaget suggests also that the young child's thoughts about and
perceptions of stimuli are 'egocentric', or self-centered. For instance
the young child interprets a stimulus according to whichever aspect of
the stimulus first drew his attention and he believes that everyone else
views the stimulus as he does. About the time the child enters first
grade, however, he is losing this egocentric orientation. This change
implies that the older child will 'decenter', or attempt to consider and
integrate all of the elements of the perceived pattern into a whole. It
further implies that the older child's thoughts and perceptions are fairly
flexible and less often dominated by that element of the stimulus that
elicited his attention initially.

Bruner et al. (1966) hypothesised that young children--even though
they do decenter--can integrate the elements of a simple, familiar
pattern only. They predict that the ability to integrate the elements of
involved patterns emerges during early adolescence.

Considering Piaget and Bruner together, the cognitive-developmental
lpproach seems to imply that children of all ages may look at all elements
Of a stimulus, but that conceptual development--the ability to integrate

Ill.of the information obtained improves with age.

11
Summary.

There are implications in several current theories for the study of
exploratory or curiosity behavior. Berlyne's theory is primarily concerned
with drive reduction. It designates the stimulus dimension,,zig., complex-
ity, novelty, surprisingness, incongruity, change, and uncertainty, that
elicit curiosity behavior. Hebb's theory suggests that simultaneously
aroused expectations disrupt behavior patterns by disrupting or cancell-
ing the underlying neurological phase sequences. Piaget's work suggests
that any of Berlyne's stimulus properties may elicit responses similar
to the exploratory behavior of the assimilation period (e.g., persistent,
repetitive responses toward an incongruous stimulus) and that observation
of this behavior is more likely among older elementary school children

than younger elementary school children.

Definitions of Incongruity

Incongruity, the stimulus variable with which this paper is con-
cerned, poses a number of quite difficult problems of definition. The
term 'incongruous' usually implies the unexpected, out-of-place, and
internally inconsistent, as determined by the perceiver's past experiences.
The theoretical definitions offered by Berlyne (1960), Hebb (1949),

Maddi (1961), Harvey (1963), and others seem to imply much the same thing.

According to Berlyne, incongruity "exists when a stimulus induces
an expectation which turns out to be disappointed by the accompanying
stimuli" (1960, pp. 24-25). The expectations are due to the “synchronous
predictive redintegrstion" (p. 25) activated by some stimulus. In other
words, for each stimulus there is a neural correlate and an associated
response; an incongruous stimulus simultaneously activates two neural

correlates, and as the cerebral neurons for each response fire (or the

12
threshold for their firing is lowered), the responses conflict so that
both cannot be carried to completion.

Berlyne (1960) further suggested that incongruity is a unique combi-
nation of the other collative properties. However, merely codifying and
quantifying the collative properties of incongruity will not yield a satis-
factory definition because, as noted above, the ordinary use of the term
implies that incongruity is also a function of the observer's previous
perceptual experience. The definition presented by Berlyne stresses
previous experience less explicitly than do the definitions offered by
Maddi (1961) and Lore (1965), or Harvey (1963).

Berlyne's definition is unwieldy for research because its use in
research would require pretraining sessions during which each subject
learned a specific response to the congruous drawings and a testing
session for observing the effects of incongruity on the learned response.
For this reason, another definition was used for the current research.

Other definitions are implied or stated by Lore, Hebb, Piaget, and
Harvey. Lore (1965) contends that incongruity is a relational concept
that can be defined only in terms of the physical properties of the
stimulus and the individual's previous experience with objects of the
same stimulus class. As indicated above, beb would define incongruity
as a stimulus property leading to the disruption of cell assemblies as
a result of a defeated expectation. Piaget (1952) indicates that the
moderate modification of a familiar pattern is the stimulus character-
istic most effective for eliciting orientation and exploration. It seems
that Piaget is referring to incongruity. Length of looking time, a de-
pendent variable in the present study, is suggested by Piaget's defini-
tion. Harvey (1963) defines incongruity as the failure of new receptor

input to match stored information--either on a short term basis, as when

13
there are rapid changes in receptor input, or on a long term basis, as
when "objects, situations, persons, and Ereviousla sequentially organised
events fail to match the inputs that come" (pp. 60-61). Because it would
be difficult to measure precisely receptor input and stored information,
Harvey's definition was not used.

Maddi (1961) offers still another definition: "An incongruous stim-
ulus contains elements which are perceived as incompatible with each
.other on the basis of prior experience with these elements” (p. 256). He
suggests also that incongruity may be directly related to complexity, and
that it is simdlar to surprisingness except that it involves a spatial
rather than a temporal dimension (p. 267). The general--as distinct
from the operational--definition of incongruity used in the current study
seems closest to Maddi's. This definition was chosen because an opera-
tional definition could be derived from it, i.e., it is possible to
determine-~with reference to specific stimuli--what combinations of stimv
ulus elements might be perceived as incompatible with each other on the

basis of past experience.

Types of incoggguity.

Though the present study is about visual incongruity, incongruity
also can be experienced with the other senses, and it can be experienced
between sense modalities. For instance, one may see a shiny surface and
expect it to feel smooth but discover that it is rough to the touch.

Epistemic, or conceptual, incongruity refers to the symbolic con-
flict arising when properties occur together after one has learned that
they occur separately. Though this would seem to be an interesting direc-
tion for research, few psychologists other than Berlyne (1954, 1962)

have studied epistemic curiosity.

14
Previous Research

Although there has been much research on curiosity in general and on the
properties 'novelty' and 'complexity', little research has been directed
to the behavioral concomitants of incongruity. Instead, incongruity has
been studied as a component of novelty in research with both animals and
humans. Therefore, the following review of research deals with studies
of novelty and complexity and attempts to relate the findings to the

study of incongruity and perceptual curiosity.

Studies with animals.

Animal studies have dealt with both object novelty and situational
novelty. For a more comprehensive review of the early animal literature
the reader is referred to Berlyne (1960), Helker (1961), and Brown (1961).
The present discussion will be limited to a review of a few selected
primate studies.

Butler (1954), for example, found that monkeys learn to discriminate
between two colors and to respond to the discriminative stimulus when
the reinforcement is a chance to peek into the laboratory, i.e., a change
in visual stimuli. The monkeys peeked more often at a changing than an
unchanging scene. Harvey (1963) interpreted the higher response rate to
the bustling laboratory as a function of the greater incongruity in the
changing laboratory scene. Actually, it seems as though the monkeys'
choice was for relatively more complexity and relatively more novelty
rather than for incongruity. The animal's cage was familiar and probably
barren, so that the laboratory, while not completely unfamiliar, was
more novel and more complex than the cage. The more 'active' laboratory,
of course, provided.mnre change froh the cage than did the quiet labora-

tory. Another significant feature of Butler's study was the endurance

15
of the operant 'peeking' response reinforced only by perceptual
curiosity.

Helker improved on Butler's by employing a naturalistic setting and
a spontaneous rather than a conditional response. Walker (1956., 1956b,
1956c) found that young chimpanzees looked at and touched the novel,
heterogeneous stimuli more than did older chimpanzees. There were several
sets of novel objects, and each presentation of a different set the
exploratory behavior, albeit extinguished on the preceding set, re-
appeared. The young chimpanzees had consistently longer approach laten-
cies to each new set of objects than did the older subjects.

Hebb (1946), too, reported that chimpanzees responded to novel
stimuli. These chimpanzees, however,did not approach novelty - they
avoided it, screamed, and shook their cages. Hebb called their responses
"fear". It is possible that the longer approach latencies of Welker's
young chimpanzees indicates mild fear also.

Both Hebb and Velker could be sure that their stimuli were novel
since they each were familiar with their subjects' perceptual histories.
However, the novelty in Hebb's stimuli lay in their being rearrangements
of the elements of familiar stimuli. Thus the dominant characteristic

of his novel stimuli appears to be incongruity.

Studies with humans.

Like the animal literature, most of the curiosity research with
human subjects centers on novelty and complexity. A few of these studies
have dealt with incongruity as a dimension of novelty or, less frequently,
of complexity. The subjects typically have been adults, nursery school
and kindergarten children, and, in a few cases, infants. One study used

elementary school age subjects. There has been no truly developmental

16

study so that age trends must be inferred from a variety of studies,
and any inferences must be tentative because of the great variety of
methods and analyses used.

loveltz. 'Host studies with humans have dealt with "relative
novelty", since it is easier to manipulate recency of stifiulus encounter
than the total perceptual history of a subject. The research of Berlyne
(1954, 1957, 1958‘, 1958b, 1962), Cantor and Cantor (1964), Clapp and
Bichorn (1965), Pielstock, Woodruff, and Bond (1964), Lintz, Starr,
and Hedinnus (1965), Harris (1965, 1957a: 1967b), Lore (1965), and
Smock and Holt (1962) has shown that children and adults tend to
"like" (Berlyne's term) novel pictures and objects and that they re-
spond more to these stimuli than to familiar stimuli. Harris (1965)
reported that children choose one damaged novel toy over two undamaged
familiar toys. 0n the other hand, the research of Hosts, Miller, and
Spitz (1963), Burgess (1957), Hagan et al. (1966), and McCall and Kagan
(1967) has not shown significant preference for novel stimuli or greater
responsiveness to novelty. The difference in these studies may lie in
the age of subjects used, kind of stimuli, etc., as we will see below.

'1, novelty and response freguency. Using exposure frequency as
a measure of curiosity, Berlyne (1957) found that college students
looked more at incongruous than at congruous pictures and more at novel
than at familiar pictures. This result has been found only inconsis-
tently with nursery school children (Clapp and Eichorn, 1967; Burgess,
1957; Lore, 1965).

2. loveltz and response duration. Berlyne (1957), Cantor and
Cantor (1964), Johnson (1967), New and New (1961), Pielstock et al.
(1964), and Burgess (1957) have used duration of looking time as a

behavioral measure of both preference and curiosity. Longer duration

17

of looking time has generally been associated with stated prefer-
ence for the collative variable. Subjects looked longer at relatively
more novel stimuli.

3, Hoveltp and physiological peactions. Lore (1965), Kagan et al.
(1966), and McCall and Hagan (1967) reported relations between stimulus
selection (as indicated by duration of looking time) and certain physio-
logical correlates of attention. General activity level and cardiac
rate decreased while skin conductance increased. Lore (1965) found a
correlation between these physiological indexes of attention and the
mean duration of looking time at incongruous and novel stimuli for a
group of middle class kindergarten children but not for a matched
group of "underpriviledged" children. It is especially in research
with infants that the effects of novelty and incongruity seem.most
inseparable. Kagan et al. (1966) found that cardiac deceleration of in-
fants occurred significantly more often within the first 3 seconds of
the initial fixation to a picture of a normal face (familiar stimulus)
than to a face with 'scrambled' features. McCall and Hagan (1967)
obtained the same results using both photographs and drawings of faces.
In these two studies the responses were interpreted as "recognitory".

It seems likely that infants would respond to a previously perceived
stimulus because some of its cues were learned. However, infants do

not have well-developed schemas for familiar objects, patterns, per-
sons, events, etc. Thus discrimination of incongruity is unlikely
because there is no standard with which to compare the unfamiliar stimu-
lus. After the infant has discriminated the particular and fixed
patterning of the elements of several stimuli, he may be more affected

by a rearrangement of the components of those stimuli.

18

4, Novelty and affective responses. Kagan et al. (1966) have
reported affective responses in studies of curiosity. They found that
4-month-old infants smile more to normal faces than to scrambled faces.
There was no evidence of differential affective responsiveness among
their subjects to novel, i.e., modified familiar (hence incongruous)
stimuli. (Hebb and Welker obtained essentially the same results with
young chimpanzees.) Piaget (1952) notes that infants often respond
effectively while observing slightly modified stimuli. The infants
show "surprise", "puzzlement", ”startle”, smiles, and laughter. The
expression changes, according to Piaget's reports, seem to follow the
initial orienting reaction and occur simultaneously with exploratory

responses.

5, _Incongpuity as a dimension of novelty., Berlyne-(1957, 1958,
1963),?using adults, and Clapp and Eichorn (1965) and Lore (1965),
using preschool and kindergarten children, found greater responsive-
ness to incongruous than to congruous stimuli. Burgess (1957), on
the other hand, found no significant preference for incongruity among
nursery school children. Berlyne (1958) controlled for "identifica-
tion time" and found that the adults' preference for incongruous
stimuli was not a function of identification time alone.

In conclusion, it seems that incongruity may not be an effective
dimension of novelty with children younger than elementary school age.
Responsiveness to incongruity does not seem to be a function of
stimulus identification alone.

Cgpplexity. Complexity is the second of the two "collative
properties" to receive considerable attention by psychologists. Com-

plexity is relatively easy to define physically by using standard

information-theoretic means, but its psychophysical component is

19

apparently closely tied to the individual subject's past experience
just as in the case of other collative variables.

1, Pattern complexity and preference. Numerous studies using
subjects within the age range of approximately 5-18 years (Berlyne,
1957, 1958, 1963; Burgess, 1957; Clapp and Eichorn, 1965; Pielstock,
Woodruff, and Bond, 1964; McCall and Hagan, 1967; Hunsinger and lessen,
1964; Hunsinger, lessen, and lessen, 1964; Smock and Holt, 1962;
Munsinger, 1965; Hunsinger and Kessen, 1966) suggest an age invariant
verbally-stated preference for stimuli of intermediate degrees of can-
plexity. However, younger children show a greater preference for more
complex forms than do adults (Plelstock et al., 1964; Munsinger and
lessen, 1964; Hunsinger, lesson and Kessen, 1964).

Using a paired-comparison design, Berlyne (1963) found that adults'
stated preference correlated negatively with duration of looking time,
i.e., more adults rated less irregular (LI) patterns as more pleasing,
but more irregular (HI) patterns as more interesting. When allowed
to choose a pattern for further viewing the adults chose the MI patterns
significantly more often than the L1 patterns.

In summary, it seems that children and adults prefer patterns of
moderate complexity. When asked to choose patterns that are more
interesting, adults choose HI patterns, i.e., patterns similar to those
‘which children prefer. When asked to choose "pleasing" patterns, adults
choose LI patterns, i.e., patterns similar to those which older children
prefer.

2, Pattern complexity and recognition. Munsinger and Kessen
(1966) found that although younger children prefer the more irregular
patterns, they remember the L1 (meaningless) patterns more readily

than the MI ones. Adults were slightly superior in the recognition

20
of previously seen complex patterns.

3, Pattern complexity and response duration. Johnson (1967)
reported a longer looking at more complex than at simple patterns by
adults and children.

4, Pattern eggplexity, preference, and intelligence. Pielstock
et al. (1964) found that "gifted children" (Binet IQ? 132) are more
responsive to pictures of greater variability than are children of
average intelligence (Binet IQ 92 - 108). It thus seems as though
intellectually gifted children prefer the HI patterns, whereas average
children prefer the L1 patterns. The responses of average children
were similar to those of older children and adults. The preferences
of the sixth-grade gifted children, however, were more similar to the
adults' preferences than to those of the young gifted children or the

young average children.

Summary

The foregoing review suggests that little is yet known about the
relation between incongruity and curiosity or about the relation be-
tween age and incengruity. Several studies have investigated the
curiosity-inducing property of incongruity as a dimension of novelty
using duration and frequency of looking time and physiological measures
as dependent variables. No study has attempted to deal with incongruity
as a dimension of complexity, though this suggestion was made by Haddi
(1963). No study has investigated the relation between age and incon-
gruity, although several studies using subjects of varying ages have

included a few incongruous pictures in the set of novel stimuli.

21

The Present Study

The present study focuses on the visual exploration of incongru-
ous pictures, with special attention directed to children's spontaneous
responses to pictures. The incongruous stimuli were outline drawings
consisting of the head of one animal and the body of either another
animal or of an inanimate object. (This operational definition thus
does not take into account each child's unique experiences, but the
study was concerned more with between-subject than within-subject com-
parisons. The current operational definition and stimuli are similar
to those used by Berlyne (1960), Lore (1965), and others.

The incongruous figures and congruous 'control' figures were shown
to kindergarten, second-grade, and fourth-grade children. These groups
were used to expand the age range employed in previous studies, most of
which have used only pre-school and kindergarten children.

The investigation was concerned with several response classes
that might indicate sensitivity to incongruity or other collative
properties in the child's everyday life. These responses are:

(1) The duration,or length, of time the subject looks at the
stimulus picture and frequency of looks at the stimulus. These
measures, as noted earlier, have been used frequently as indexes of
perceptual curiosity. Looking time frequency was defined as the
number of looks at each stimulus, and looking time duration as the
length of time spent looking at each stimulus.

Looking time seems to be the obvious and natural index of stimulus
selection. However, because it is usually a demand characteristic of
the perceptual situation, it probably has limited value when used alone.

It should be more meaningful when combined with other measures.

22

(2) The number of times §,changes his expression while viewing
a picture, the incidence of an expression change to the congruous and
incongruous pictures, and the duration, or length, of all expression
changes to each picture.

Casual observation of adults and children indicated that smiles,
laughter, frowns, widened eyes, and comments were reliably associated
with observation of the stimulus. Expression change has been reported
as a dependent variable by very few psychologists, although an early
exploratory study by Swan (1938) indicated that expression change was a
potentially fruitful dependent variable with nursery school children.
Swan reported that the children were very expressive: frowns, chuckle-
laughs, and talking-to-self occurred at least once for 751 of the
children in a 90 minute free play situation, and all children smiled
at least once. Other facial expressive behavior included laughing,
crying, lifting eyebrows, pursing lips, and sucking. Further, it was
noted that children.who talked more showed fewer silent facial expres-
sions. Charlesworth (1968) suggested that facial expressiveness may
provide information about the subject's cognitive activity. An ex-
pression change measure may disclose age changes in the perception of
incongruity insofar as that perception depends on the complexity of
the subject's schemas and his level of cognitive development.

Psychologists have not used affective responses as dependent
variables because they can be inhibited and/or faked, and because they
are thought to be very dependent upon the subject's past experiences,
personality, and age. Though these reasons may be valid for adoles-
cents and adults, the evidence suggests that affect is expressed rela-
tively openly by children in the age range used in the current study

(Miller, 1960).

23

(3) The name assigned to each incongruous stimulus. 'Hany sub-
jects in a pilot study spontaneously named the figures while viewing
them. Some of the children named just one component of the incongruous
figures, while other children named both components or tried to make up
a new name using the names of both figures. Labeling therefore was used
to help decide whether the differential responsiveness to the figures
could be interpreted as perception of incongruity or whether it might
be a visual habit.

Psychologists frequently have attempted - but with little success -
to relate intelligence (IQ scores) to responsiveness to visual stimuli.
Curiosity indeed might be expected to correlate with the scores on some
subtests of the Stanford-Binet, such as 'picture absurdities', but not
with others such as 'arithmetic'. The low correlations found between
IQ scores and responses to visual stimuli therefore should be of no
surprise. A pilot study suggested that labeling was a direct and reliable
indication of the child's ability to recognize and communicate informa-
tion about pictures and as such could be a useful measure of cognitive
development.

Since older children undoubtably have had more experience with
animal pictures than have younger children, they ought to be able to
label the pictures more rapidly and specifically.' However, the current
research did not study labeling per se but rather sought to use spon-
taneously emitted labels as an indication of perceived incongruity.
Therefore age differences in labeling ability need not necessarily be
considered a drawback. A pilot study indicated that young children could
assign an animal class name, e.g., 'bird' and 'dog', to any component
of an incongruous figure that they could not name exactly, e.g.,

'flamingo', 'wolf'. It was felt that such naming was a sufficient

24
indication of whether or not the child perceived a figure with two in-

compatible components.

Predictions.

The predictions below stem from Berlyne's theory of the effects of
the collative properties, and from the research cited above - especially
the studies of Berlyne, Burgess, and Lore.

(1) The mean duration of looking time to incongruous figures will
exceed the mean duration looking time to congruous figures.

(2) The mean duration of looking time to incongruous pictures will
be longest for kindergarteners, intermediate for second-graders, and
shortest for fourth-graders; Alternatively, this same ordering is ex-
pected when the statistic of interest is proportion of total looking
time spent looking at the incongruous figures because younger children
simply may look longer at everything.

(3) The mean duration of expression change to incongruous
figures will exceed the mean duration of expression change to congruous
figures.

(4) The mean duration of expression change for kindergarteners
and second-graders will exceed the mean duration of expression change
for fourth-graders.

(5) The mean number of expression changes to incongruous pictures
will exceed the mean number of expression changes to congruous pictures.
(6) There will be a direct relation between acknowledgment of
‘both components of the incongruous figures and grade in school, though
it is expected that the difference will be larger between kindergarten

and second-grade than between second and fourth grade.

METHOD

Subjects.
The subjects were 60 children (27 boys and 33 girls) enrolled at

the Spartan Village School on the campus of Michigan State University.
The sample included 20 kindergarteners (5 years 3 months-~6 years 10
months), 20 second-graders (6:2--8:5), and 20 fourth-graders (9:2--

11:7). There were 9 boys and 11 girls from each grade.

Mhterials.

The stimuli were designed using picture books, coloring books,
etc., as a guide to what should be familiar for children of the age
range studied. Thirty pictures were designed: l5 congruous (C) and
15 incongruous (1).. The C stimuli were ditto copies of animal draw-
ings. The I stimuli were ditto pictures of combinations of two con-
gruous animals such that the completed figure had the head of one
animal and the body of another. Several figures were combinations of
an animal's head and the body of an airplane or a car. (Copies of
the congruous and incongruous pictures used are included in Appendix A.)

There was no control for complexity as measured by number of inde-
pendent turns, number of sides, etc. CMunsinger and Kessen, 1964).
However, the pictures were designed with minimum internal detail.
Gibson (1963) reports that simple drawings, such as cartoons, facili-
tate children's perceptual performance in picture identification by
"eliminating noise from the stimulus" (p. 189).

The congruous pictures were used to control for novelty, i.e., if
the congruous drawings were novel, Sp would probably respond similarly
to both novelty and incongruity making it impossible to determine which
had been discriminated.

25

26

It was expected that the older children would be more faniliar,
because of longer experience, with the kinds of pictures typically
shown in picture books (e.g., with the various zoo and farm animals).
Nevertheless most of the kindergarten children, after 7 months in
school, were able to identify the congruous animals and to handle the
books in which they were bound with skill sufficient to that needed in
the task. It is very unlikely that any of the subjects would have seen
the particular incongruous figures used in this study.

Pour identical sets of the original 30 figures were bound into 11
inch x 18 inch folders. Four C - I randomizations were carried out
using a revised Cellerman sequence (Gellerman, 1933). The C - I
randomizations were performed in blocks of ten (3 blocks for each
book) with the following restrictions: (a.) there were five C and five
I figures in each block; (b.) there were no more than three C or three
I figures in succession; and (c.) the first figures in each book were
one C and one I. The figures were assigned numbers (C - l - 15, I -

16 - 30), and then, according to the above criteria, were arranged with-
in each book so that each book contained the same pictures but in differ-
ent orders. The orders used are presented in Table 1, Appendix B.

Because younger children might be less adept in handling the mater-
ials, thick pages were used to permit easy turning by even the youngest
subjects. (Pages that are difficult to turn could draw the child's
attention from the pictures to the duty of obeying g]: instructions to
turn only one page at a time.)

The picture book was placed on a bookrest which was fastened to
the table with masking tape. The back of the rest slanted forward so
that the book would rest on the ledge and back of the bookrest. The

bookrest was approximately 18 inches x 12 inches and was positioned

27

so as to permit easy view of the pictures for even the youngest.§s.

Procedure.

Ten days prior to the starting date, §,visited each of the parti-
cipating classrooms for 30 minutes in order to meet and talk to the
children. At this time §.described the children's role as follows:

I guess just about everyone likes to look at pictures;

I know that I do, and I bet that you do too. There are

lots of different kinds of pictures and we see them in

many places. There are pictures in books and magazines,

pictures we hang on the wall, and moving pictures. I have

some pictures that I have drawn and put into a book for you

to look at. Some of the pictures you have seen before and

some will be new and different.

It is important that you don't tell anyone else about the

pictures once you have seen them. We want them to be a

surprise for everyone. Do you think you can keep it a

secret: I think that you will enjoy looking at the pictures.

0n the first day §.escorted the first child from.his classroom to
the test room, an 11 x 8 foot spare room in the school. The door of
the room contained a one-way mirror. The §_sat before the book stand.
Sometimes the younger g; had to be propped up to enable clear view of
the pictures and easy manipulation of the pages. The table (3 feet
high) for the bookrest was placed in front of the one-way mirror such
that g was 7 feet from the mirror. The gs were seated in the adjoining
room 1% feet behind the mirror. In that room they were hidden behind
a screen so that §.could not see them as he entered the test room.

The closed book lay on the bookrest and remained balanced as‘g
turned the pages. The §_spoke the following instructions from.memory:

Remember when I came to your room and told you about the

picture books that I had? (§_reminded §s who had forgotten.)

Well, this is one of them; the first page is empty but each

page after that has a picture on it. You are to turn the

pages one at a time, like this (§_lifted the blank page by

the bottom corner) and look at each picture for as long as you

want. Then turn the page and look at the next picture for as
long as you want. Remember to turn just one page at a time.

23

Any questions: Wait until I sit down over here and
then you may begin. 0k, go ahead.

The 1; then moved to a corner of the room approximately 6 feet from _S_.
From this position she could see _S_'s eyes clearly.

After g looked at the last picture g said,

Now let's go back to the first page with pictures on it.

I would like you to look at each picture again and name

it for me so that I can write it down here on this paper.
If g paused or asked about a picture, g said "You give it a name. Tell
me what it looks like." The 1: did not suggest that more than one name
might be "correct", or that there was a "correct" name.

After the naming, g returned to his classroom, brought the next

g to the test room, and then rejoined his class.

Apparatus.

The recording apparatus was a four channel Rustrak event recorder,
operated simultaneously by E and two observers, 21 and Q2. The 1; was in
the room with g, but Q's hand control was concealed and operated silent-
ly. A fan in the test toom masked outside noises. The observers and
the event recorder were in the adjacent room and were never seen or
heard by g.

After the instructions 91 started the event recorder. The‘g
depressed the button on her hand control when §,began to look at the
picture and released it whenever §,looked away from the picture. The
21 recorded each page turn.with two 'bltps' of a second control, and
with the third hand control recorded the onset and duration of §fs
changes of expression. The inrecorded onset, duration, and offset of
§fs looks at §,with a fourth hand control. All four controls were wired

to the single recorder for simultaneous recording of the various events.

29
Dependent vapiables.

Duration of looking time was the length of time §_looked at the
picture. It was computed by subtracting the duration of looks away
from the drawing, e.g., looks at g, looks at the edge of the page, looks
around the room, from the total picture presentation time. Picture pre-
sentation time was that period of time between page turns. A page turn
began when the overlying page was raised to approximate vertical with
the next page, and the page turn ended when that page was raised to
approximate vertical with the next page.

An expression change was defined as any change in expression.while
§_observed each drawing. Expressions such as the following were re-
corded: smiles, frowns, laughs, raised eyebrows, widened eyes, opened
mouth, pursed lips, creased forehead.

Each §ywas asked to name all pictures. The child was presumed
to have perceived both components of the incongruous figure if he
labeled both correctly, or if he mislabeled one (i.e., if his name
for one component was in fact incorrect but nevertheless was different
from.the name for the first component, e.g., duck with dog's head in-
stead of duck with wolf's head).

geliabilitz of dependent variables. Prior to the testing, the
reliability, or percent agreement, for each dependent variable was

assessed using the following formula:

total number of overla in recordin s p 100

total number of separate recordings
The reliability for looks at §_was 83.1; for expression change, 76.5;

for non-orientation to the stimulus, 96.8.

30
The apperimental setting.

In the most general sense, this research seeks to identify the
natural attractors of children's attention and to determining the be-
havioral indexes of that attention. To accomplish these aims, it is
important that feelings of self-consciousness, anxiety, and fear be
minhmised. Primarily for these reasons, certain considerations were
observed in the design of the current study. The children were tested
in a familiar room in their own school. The table and chairs in the
roomnwere similar to those used in all other rooms in the school.

In a pilot study, 4-to 6-year old children enthusiastically looked
at some c and I pictures when asked to do so by a friendly adult, and
everyday observation indicates that most elementary school children
are familiar with and like to look at picture books. The stimuli used
in this study, therefore, were bound into books designed to resemble
picture books.

The demand characteristics of the situation were minimal. General
experience with children indicated that the fewer specific demands
made upon the child, the more spontaneous his behavior. The child in
the current study was almost completely in control of the situation.

He simply was asked to do what he would do normally--to look at a picture
book. He had two rules to remember while looking at the book, both

rules to insure uniform behavior across all children and to permit re-
cording of his behavior. Few children had trouble turning only one

page at a time.

In much previous research, the stimuli have been presented either
by §.or §.in a tachistoscope or (timed) slide projector. But the use of
mechanical gadgets with young children can be criticised inasmuch as

the child's interest in operating or controlling g}: operation of the

31
mechanical device may bias the results. (Indeed, this may explain
why the CD5, or "culturally deprived," subjects in Lore's study did not
respond differentially to incongruity and asymmetry. It may also ex-
plain why Burgess did not obtain significant differences in looking
times for his 5-year-old subjects.) Also, young §fs behavior may re-
flect boredom or satiation with a gadget rather than responsiveness
to the experimental stimuli. No one seems to have reported any sort
of control for the effects of the presentation apparatus. The absence
of any mechanical gadgets in the test room in the current study makes

such controls unnecessary.

RESULTS

Duration of looking time.

Duration of looking time (length of time1§ looked at a picture)
was calculated as total length of time (in seconds) the picture was
presented minus total duration (in seconds) of all looks away from the
picture. The scores for all g; were summed and divided into 3 blocks
of 10 trials, each block containing 5 incongruous pictures and 5 con-
gruous pictures. An analysis of variance (Table 1) with congruity-
incongruity and trial block as 'within' variables and grade as 'between'
variable disclosed significant effects for the congruity-incongruity
dimension (3, - 17.80, if - 2/114, p (.01). Eons-way analysis of
variance for pairs of trial blocks disclosed a significant incongruity
x trial block interaction for trial blocks I and III only (£,- 7.75,
d_f. I 1/118, p (.Olg No other main effects or interactions were sig-
nificant.

The mean duration of looking time across trial blocks was 7.99
seconds @ - 6.87) to the incongruous figures and 7.71 seconds Q .
4.85) to the congruous figures. Table 2 presents these totals in
addition to subtotals for each grade in school and each trial block.

As can be seen, across all trial blocks and all grades the mean duration
of looking time was longer to incongruous figures than to congruous
figures.

The foregoing analyses were concerned only with group effects.

The incongruity-congruity effects also may be viewed in terms of the
scores of individual subjects. Figure 1 presents, by trial block, the
total number of subjects in each grade level who looked longer at the

incongruous than at the congruous figures. As can be seen, an equal

32

33

TABLE 1

SUMMARY 0! ANALYSIS OF VARIANCE OF DURATION OF LOOKING

TIME TO INCONGRUOUS AND COHQUOUS ANIMAL DRAWIME

Source g; ‘ SS HS 3 p
Between 59 245,339.070
Grade (C) 2 '2,l76.3l 1,088.155 (1.00
Errorh 57 243,162.76 4,266.018
Within 300 96,700.842 . '
Trial Block (B) 2 15,321.695 7.660.548 17.8066 (0.01
C‘x B 4 1,585.335 398.385 (1.00
Error 114 49,043.70 30.20
Congruity (A) 1 3,378.311 3,378.311 19.5235 ‘<0.01
0 x A 2 121.582 60.791 (1.00
A x Errorb 57 9.863.136 173.657
8 x A 2 379.097 189.548 1.2814 (0.05
C x B x A 4 > 205.718 51.480 4(1.00
Al x Errorb 114 16,862.865 147.919

Total 359 342,099.9125

34

TABLE 2

MEAN DURATION OF LOOKING TIME TO INCONGRUOUS AND

CONGRUOUS FIGURES AS A FUNCTION OF GRADE IN SCHOOL

TRIAL BLOCK
I i-
E-
II '1?-
.82-
III i-
Q-
Total :-
s_D-

AND TRIAL BLOCK

GRADE IN SCHOOL

Kindergarten
Incon, Con,

10.26 8.47
(7.16) (4.82)

6.74 6.27
(3.30) (2.64)

5.27 4.80
(2.15) (2.17)

Second

Inpon, Con,

10.41 8.67
(3.85) (4.09)

8.59 7.18
(2.83) (2.28)

6.66 5.91
(1.90) (1.57)

 

Fourth

Incon, Con,

, 8.56 7.02
(1.87) (1.19)

7.71 5.71
(1.98) (0.98)

6.26 5.41
(1.22) (1.09)

 

Total

Incon on

9.74 8.05
(9.60) (6.96)

7.68 6.38
(5.57) (4.23)

6.06 3.65
(5.37) (3.37)

 

7.42 6.51
(4.20) (3.21)

8.55 7.25
(2.86) (2.64)

7.51 6.04
(1.69) (1.09)

 

7.99 7.71
(6.87) (4.85)

35
number of subjects in each grade looked at the incongruous figures
longer than at the congruous figures in trial block I. In trial blocks
II and III, however, the number of kindergarteners and second-graders
looking longer at the incongruous figures remained constant or de-
creased, while the number of fourth-graders increased. Across all 3
trial blocks, 20 of the 60 subjects always looked longer at the incon-
gruous figures than at the congruous figures.

Table 3 presents the frequency distribution for the 8 possible
patterns of longer duration of looking time across trial blocks, i.e.,
whether the longer looking time per trial was to the incongruous
figures (I) or to the congruous figures (0). Five kindergarteners,

6 second-graders, and 1 fourth-grader looked longer at the congruous
figures in each of the 3 trial blocks. Finally 12 kindergarteners,

15 second-graders, and 17 fourth-graders looked at the incongruous
figures longer than at the congruous figures in 2/3 of the trial blocks.

Still another way to examine the duration of looking time data is
to consider the percent of total looking time to the incongruous figures;
Table 4 presents these percentages. The mean percent of total looking
time at incongruous figures was 541. As can be seen, across all trial
blocks and across all levels in school, more than 501 of the total
looking time was to the incongruous figures. For kindergarten and second-
grade, percentage of total looking time to the incongruous figures de-
creased regularly but slightly across trial blocks. The distribution

of percentages for fourth-graders was irregular.

ngation of oppression change.

Across all §p the mean duration of expression change to incon-

gruous figures was 2.35 seconds (§Q_- 0.85), to congruous figures,

Number of subjects

Figure l.

36

Number of subjects (per trial
block) with longer looking times
to incongruous than to congruous

figures.

 

   

(v

 

-chance
level

 

 

Second-grade

- O - o o - O 0 so

Kindergarten

II III

Trial Block

 

37

TABLE 3
FREQUENCY DISTRIBUTION OF THE 8 POSSIBLE

PATTERNS OF RESPONSE FOR EACH GRADE LEVEL

8 Possible Patterns of Longer
Duration of Looking Time to

Gongruous (C) or Incongruous (I) Second Fourth
Figures per Trial Block : Kindergarten Grade Grade
1. I I I Hi 5 6 9
2. I I C 5 4 3
3. I C I 1 3 2
4. C I I 1 2 3
5. I C C 3 1 0
6. C I C 0 2 0
7. C C I 2 0 2
8. C C C 3 1 1
special

0 c-I C j;

 

Total 20 20 20

38

TABLE 4

PERCENT OF TOTAL LOOKING TINE TO INCONGRUOUS
FIGURES AS A FUNCTION OF TRIAL BLOCK

AND GRADE IN SCHOOL

 

 

Grade Trial Block
in school I II III Total
Kindergarten 55% 521 521 531
Second 55% 542 5 31 541
Fourth 551 571 541 542
Total 552 54% 531 54%

 

 

39

0.70 (§2_- 0.59). Analysis of variance (Table 5) indicated that
this difference was significant (§.- 19.87, d£,- 1/240, p_ .0005).
The analysis also disclosed a significant book effect (E,- 2.81,
g;,- 3/48, p_<{.05), but no effects of grade, trial block, or any
combination of grade, book, trial block, and congruity.

As shown in Figure 2 the mean duration of expression change was
longer to the incongruous pictures than to the congruous pictures in
each book, though, for reasons as yet not understood, Book 2 elicited

significantly longer expression changes than did the rest.

Freguencz of pgpression change.

Frequency of expression change is the total number of expression
changes to each picture. Across all gs the mean frequency of ex-
pression change to incongruous pictures was 1.68 (§Q,I 0.43), and to
congruous pictures, 1.45 (fig,- 0.45). This difference, however, was
not significant Q - (1.0, g; - 1/240, p (0.70).

-The failure of the frequency of expression change measure to
interact significantly with the congruity-incongruity dimension may be
due partially to the significant interaction of congruity and duration
of expression change. Observation of g; disclosed that some expression
changes were much stronger than others. For instance, some expressions
were 'weak', e.g., turning up slightly the corners of the lips. Others
were more intense, e.g., laughter. The weak expression changes were
shorter than the more intense changes. For the most part the expression
changes made to congruous pictures were short (and week), while those
made to incongruous pictures were long and intense. Thus several
weak expression changes could be made in the time required for one

intense expression change so that the frequency of expression change

40

TABLE 5

SMRT 0! ANALYSIS OF VARIANCE OP MEAN DURATION OF EXPRESSION

CHANGE TO INCONQUOUS AND CONGRUOUS ANIMAL DRAWINGS

Source 51$ SS HS 3 p
Between 59
Book (A) 3 621.646 207.215 2.813 (0.05
Grade (C) 2 274.801 137.403 1.865
A x C 6 394.466 60.744 (1.00
Error 48 3,536.431 73.676
Within 300 ' ’
Block (B) 2 2.705 1.352 (1.00
Congruity (D) 1 390.542 390.543 19.872 0.0005
B x D 2 38.492 19.246 (1.00
A x B 6 160.061 26.677 1.357
A x'B x D 6 20.803 3.467 (1.00
B x C 4 61.324 15.331 (1.00
B x C x D 4 35.944 8.986 (1.00
A x D 3 43.857 14.619 (1.00
C x D 2 21.051 10.526 (1.00
A x B x c 12 237.894 19.824 1.009
A x C x D 6 152.901 25.484 1.297
A x B x G 12
x D 69.839 5.820 (1.00
Error 240 4 ,716.570 19.652

Total 359 10, 779 .330

change

VB
..

..
.A.

”fa
J»

Mean du

tion of expressio

[\7

figure 2.

41

Mean duration of expression change as a

function of book.

 

 

 

 

‘.

i

F- e

i [:1 lecongruity
“'" a Contgrui tux
__ H

T

 

 

 

 

 

 

\\\\X\\l

 

‘

 

F.)

Look

 

 

42

effect might have been attenuated.

Incidence of first expression-change. For these reasons a new
statistic, occurrence of the first-expression-change, was employed.
Only the first-expression-change to a picture was counted; all others
were disregarded. The computation yielded 270 reactions on a total of
900 trials on which incongruous pictures appeared (302), and 85 reac-
tions on the 900 trials with congruous pictures (91). A test of
correlated proportions disclosed that this difference was significant
(p <f.001).

These data also may be considered in terms of the number of sub-
jects whose incidence of first-expression-change to all 15 incongruous
pictures equaled or exceeded the incidence of first-expression-change
to all 15 congruous pictures (incongruity2501 >congruity). There
were 17 such gs. Figure 3 depicts these scores by grade in school.

As can be seen, more than 752 of the gs of each grade level responded

more to incongruity than to congruity.

naming of incongpuous pictures.

No subject failed to give the correct name, or at least the correct
class name, to the animals or objects in the congruous pictures. For
example, the flamingo was called a 'bird' or given the name of another
bird, such as 'stork', or 'ostrich'. A subject received a score for
'incongruity acknowledgement' if two components were named for 10 or
more of the 15 incongruous figures. For the incongruous pictures, the
majority of the children either consistently acknowledged only one come
ponent in each picture, or consistently acknowledged both components
in each drawing. Across all subjects, 761 of the incongruous figures

(681 of 900) were 'named' correctly. An analysis of variance by grade

Number of children

20

18

',\
1‘

43

rigure 3. Number of subjects whose incidence of first-

exnression-change to incongruous figures

equaled or exceeded the incidence of first-

expression-change to congruous figures.

 

 

 

 

 

 

 

 

 

 

 

 

 

_ 0
——V [:1 Incongruity
-- 1— aCongruity
L.—
chance
-- - i ~+~ -~mi-~
L__
h—
Kinder- Second- Fourth-
garten grade grade

Grade in school

 

44
disclosed no signiiicsnt effects (§,- 2.877, g;_- 2/57, p_:’.05),
although the differences were in the predicted direction, i.e., the
frequency of two-coupenent responses increased directly with grade in
school.

Figure 4 sun-arises the age scores for 'incongruity acknowledge-
Ient'. The mean nnlber of components named for the incongruous figures
was 1.56, 1.72, and 1.93 by the kindergarteners, second-graders, end
fourth-grsders, respectively. Xindergerteners failed to none 10 of
the 300 incongruous pictures (32); second-graders, 3/300 (ll); end

fourth-graders, 2/300 (0.91).

Number of children

20

18

1h

12

10

CI)

45

 

Figure A. Number of children in each grade
acknowledging one or both components of
the incongruous pictures.

gllOne *

**“ .4¢ name

Both
P" names
_

....... -- ..- .-....chance..

level

 

 

 

\\\\\\\§J

 

 

 

 

 

2

 

 

 

 

 

 

Kinder- decond- Fourth-

garten grade grade

Grade in school

 

DISCUSSION

The results indicate that both duration of expression change and
duration of looking tile are related to stiwulus incongruity. Frequency
of expression change, however, was not clearly associated with incon-
gruity, though there were more expression changes to the incongruous
than to the congruous pictures. The fourth dependent variable, naldng,
indicated an increasing likelihood with age for §s to nawe both coupo-

nents, though the implied interactions were not significant.

Looking tile.
Incongruity. The prediction that looking tile would be longer to

incongruous than to congruous pictures was supported (2 < 0.01). These
results are consistent with other data in that longer looking tilcs are
generally associated with collative variables.

Trial block. The duration of looking tine weasure was clearly
associated with trial block (2 < 0.01). In general, the lean duration
of looking tile (EDIT) decreased across trial blocks Q< 0.01). Fur-
that analysis revealed that the only significant difference was between
trial blocks I and III. These results suggest that the children prob-
ably began to tire of the task as they proceeded through the book. It
also nay indicate a 'response set' toward the situation and pictures,
i.e., the children way have learned, in general, what to expect and con-
sequently adopted a response pattern for the situation.

Although the trial block x incongruity interaction was not signi-
ficant, it can be seen that, across trial blocks, the response gradient
for congruity dropped off lore sharply and steeply than did the gradient

for incongruity. If this decrease reflects fatigue or boredom with the

46

47
task, it may be that the children tired of the congruous pictures
faster than the incongruous pictures.

Grade in school. The grade effect was not significant (1,- 1.86,
g ) 0.17). However, inspection of the mean duration of looking tines
(HDLT), by grade in school, suggests some previously unknown differences
among the children in the elementary school grades.

Berlyne (1960) and others have suggested that duration of looking
time indicates responsiveness to incongruity. But to respond to incon-
gruity presupposes recognition of congruity, or the existence of a
scheme or standard for congruity, against which any incongruous stimu-
lus would be compared. Therefore it was predicted that those children
with well-developed schemes for the congruous figures (in this study,
presumably the fourth-graders) would have relatively shorter looking
times to the congruous figures than would children with less well-
developed schemes for congruous pictures (presumably kindergarteners
and second-graders). The data lend some support to this prediction.
The fourth-graders' mean duration of looking times to congruous figures
as well as to all figures were shorter than were the same mean duration
of looking times (anTs) of the second-graders' and the kindergarteners'.
The kindergarteners had the shortest HDDT to incongruous figures.
Whichever HDLTs are compared (total HDET, EDIT to congruous figures,
or HDLT to incongruous figures) there is greater similarity between the
fourth-graders and the kindergarteners than between the fourth-graders
and the second-graders. If duration of looking time indicates recogni-
tion of incongruity, then the kindergarteners apparently were more
capable of recognizing incongruity than were the second-graders. This,
however, would imply the untenable conclusion that the kindergartener's

schemes for congruous figures were more developed than were the schemas

48

of the second-graders. The conclusion is untenable because schemas
become more complex with age. The data may suggest that schemes are
developed early so that what seems to be further development is pri-
nnrily the elaboration of the basic concept. A class, for example, may
be subdivided into dogs, cats, etc. and then subdivided further. Finer
distinctions would require additional criteria and more refined dis-
criminations. Younger children may know the basic concepts'well enough
to classify an animal if the definiens of that class (e.g., wings-for
birds) are present in the stimulus. Second-graders probably have more
criteria to apply to a stimulus than do kindergarteners. Older children
probably have still more subclasses. Second-graders may look longer at
a figure to detersine to which of the subclasses it belongs--or whether
it belongs to any. Fourth-graders, on the other hand, have had more
experience applying the class criteria to stimuli and may not need to
look at figures long to classify them. I

Some of the children (not only second-graders) seemed to be trying
to identify the fitures. They asked such questions as ”What's this?”;
"Is this a rabbit?"; and others made such comments as "I.wonder what this
is?“; "This is a funny picture."; "I've never seen this before.". The
comments apparently were spontaneous and were not encouraged by‘g
either before or during the test session. Though Berlyne (1958) re-
ported that ”figure identification" is not a significant variable for
adult subjects. This suggests that further studies are needed to de-
termine whether "figure identification" is a variable that west be con-
trolled in perceptual studies with children.

Though the HDLTs of kindergarteners and fourth-graders were
similar, a comparison of the standard deviations for grades indicates

greater simdlarity between second-graders (2.74) and kindergarteners

49

(3.71) than of either grade with the fourth-graders (1.39). The
great variability in the two lower grades might influence the sample
means sufficiently to produce the above results. Indeed, one or two
second-graders responded very differently from their peers, and be-
cause the sample size per grade was small their responses skewed the
distribution so that the second-grade data may not accurately represent
a second-grade response pattern. More gs at each grade level are needed.

The raw data suggest greater variability among kindergarten and
second-grade boys than among their female peers. Simdlar'observations
have been reported (Anastasi, 1958, pp. 456-458) on intelligence scores
for young boys. The greater variability of the boys adores may be the
major cause of the skewed distributions for these-grades. Smock and
Holt (1962) found first-grade boys significantly more responsive to
incongruity than first-grade girls; indeed, girls did not respond
differentially to incongruity. The current results suggest that girls
may respond more to incongruity than to congruity. An analysis of the
current results by sex might disclose a significantly longer response
to incongruous pictures, but Smock and Holt's findings would suggest

that the difference would not be significant.

flaming.

The results lend some support to the hypothesis of a direct re-
lation between grade in school and acknowledgment of both components
of the incongruous figures. The fourth-graders acknowledged both
components more often than did the second-graders, and the seconds -
graders acknowledged both components more often than did the kinder-
garteners. Though these differences were not significant (§_- 2.88,

0.05 < p < 0.10) they suggest nonetheless that naming may be useful

50
in sheilar research on incongruity.

The usual index of ability to recognise and identify what is
familiar or complex has been the IQ score. The use of the IQ is pre-
dicated, therefore, on the assumption that whatever the test measures
is related to visual responsiveness to the collative properties of
stimuli. However, the IQ is not the only possible index of this ability.
The results of the current study suggest that naming is a direct indi-
cation of the child's ability to recognise what is familiar and to
communicate what he perceives. Naming also may be directly related
to what is being investigated in the current study, i.e., the child's
sensitivity to incongruity. The child who labels, or acknowledges,
both components of the figure indicates that he perceives at once ele-
ments that do not usually occur together. (If they occurred together,
he would so indicate by using the single appropriate label.) The use
of a double label thus suggests that the observer senses the incon-
gruity. In the current study, incongruity thus would seem to have been
acknowledged most consistently by the fourth-graders.

The fourth-graders, as well as a majority of the second-graders,
are within the age range (7 - 11 years) suggested for the developmental
period which Piaget (1952) has called 'concrete operations', and the
kindergarteners are within the 'preoperational' age range (5 - 7 years).
According to Piaget, the preoperational child tends to 'center' on one
aspect of the stimulus and, having centered, tends to resist change.

On the other hand, the concrete operational child tends to 'decenter'
and shows conceptual and perceptual flexibility, mobility, and reversi-
bility. These latter characteristics should enable the child to recog-

nise and acknowledge all of the elements that he perceives even though

51

he may not yet be able to integrate the information.

Sixty-five percent of the kindergarteners in the current study
acknowledged both components of the incongruous drawings. In terms of
Piaget's cognitive stage theory these children might be said to be
'decentering' (i.e., are into the period of concrete operations) while
the remaining 451 are still 'centering'. One-fourth of the second-
graders (two 8-year olds, two 7-year olds, and one 6-year old) behaved
as if they too might be centering insofar as they acknowledged only.one
component of the incongruous figures. One fourth-grader, 11 years old,
also responded to the incongruous pictures sometimes by giving one name,
but more often by not naming the picture at all.

The subjects could be classified as preoperational or concrete
operational only on the basis of age, and age is an imperfect criterion
for determining cognitive-developmental stage. On the basis of this
classification, the results suggest that, in general, the children who
are within the concerete operations age range acknowledged both compo-
nents of the incongruous drawings more frequently than did children not
within this age range. It must be added that this interpretation is
tentative because no independent measure of centering was obtained.

Such an independent measurement marks a profitable direction for further
research.

Another possible interpretation of the naming data is that some of
the second-grade and fourth-grade subjects had little contact with
picture books of animals prior to the experiment. But such an interpre-
tation is unlikely because these children lived in a university community
and attended school and undoubtably had seen many picture books of

‘nml‘ e

A third possible interpretation of the differences within and across

52

age groups is in terms of 'personality'. The subject might defend
himself against perceiving the incompatible elements by acknowledging
only one in each incongruous picture. The pictures were not designed
to be noxious or repulsive, but if any child perceived them in this way,
he might have spent less time looking at them. This explanation is not
supported by the looking time data. Repulsion or avoidance could be
expressed by crying, refusal to look at the book, fearful expressions,
or comments. The expression change data disclosed no such fearful
responses.

Although an interpretation in terms of centering seems more plausi-
ble than an interpretation in terms of deprivation or personality, the

current findings can only suggest the direction of future work.

Pgeguency of expression change.

It was expected that the school-age child would have well-
developed schemes for animals. It was expected also that the modifica-
tion of a familiar figure would elicit affective responses in the form
of expression changes. The prediction that expression change would be
associated more often with an incongruous than a congruous figure was
not supported, though the differences were in the predicted direction.

Across all grades, there were significantly more incidences of
first-expression-change to incongruous figures than to congruous figures.
Furthermore there appeared to be a slight but direct relation between
grade in school and the occurrence of an expression change to incongru-
ous figures. The greater responsiveness of fourth-graders suggests
that this measure is a useful one for research with elementary school
children. Indeed, expression change seemed to be gradually more discrim-

inating as age increased.

53
Duration of expression change.

The present study supports the prediction that longer expression
changes reliably indicate perception of and/or recognition of stimulus
incongruity. Some examples of the children's expression changes are
as follows: Subject 7: pursed her lips, held her mouth open and chin
slack; Subject 18: did not smile, screwed up his whole face, raised
his eyebrows, usually said the name of the picture as a question;
Subject 6 (3/25): did not smile, gritted her teeth and narrowed her
eyes until they were slit-like. There was, obviously, a wide range of
facial expressions and, likewise, a wide range in the duration of the
expression changes.

The association between duration of expression change and congruity
was highly significant (p_<:.0005). Longer expression changes, e.g.,
laughs, broad smiles, and comments, were reliably associated with child-
ren's perception of incongruity, suggesting that duration of expression
change could be used in further studies of the perception of collative
variables.

Grade effects. It was expected that the kindergarten and second-
grade children would have somewhat longer expressions because of their
greater 'openness' of expression than would the fourth-grade children.
However, duration of expression change was not significantly associated
with grade in school. The results instead suggest that the older sub-
jects may have longer expression changes. In retrospect, this finding
is not surprising, since longer expression changes were associated with
incongruity, and it was expected that the older children perceived incon-

gruity more readily than did the younger children.

54

Book effects.‘ Duration of expression change was significantly
related to book (a (0.05). Across all books, duration of expression
change was greater to incongruous than to congruous figures. Longer
expressions, however, seemed to be associated with Book 2 than.with any
other book. The reason for this association is unknown.

Before being divided into sets of 10 for the 3 trial blocks, the
pictures were randomly ordered within books following the stipulations
mentioned above under Method. Perhaps pictures should have been counter-
balanced by trial block (the same pictures occurring within a trial

block) with the trial blocks arranged in all possible orders.

In conclusion, it appears that duration of expression changes and
duration of looking time are the most reliable indexes of perception
and recognition of incongruity. Frequency of expression change poses
difficulties of definition, and it is not clearly associated with incon-
gruity. Naming, although not so reliable an index of incongruity as
the two duration measures, suggests some interesting grade differences
in the perception of incongruity.

The results suggest that duration of expression change is a poten-
tially fruitful dependent measure for future research on collative
variables. The results support much previous research insofar as

duration of looking time was associated with incongruity.

REFERENCES

Berlyne, D. E. An experimental study of human curiosity. British
Journal of Psychology, 1954(a), 3;, 256-265.

Berlyne, D.E. Theory of human curiosity. British Journal of
Psychology, 1954(b), 3;, 180-191.

Berlyne, D. E. Conflict and information therapy variables as deter-
minants of human perceptual curiosity. Journal of Experimental
Psychology, 1957, pg, 399-404.

Berlyne, D. E. The influence of albedo and complexity of stimuli
on visual fixation in the human infant. British Journal of
Psychology, 1958, $2, 315-318.

Berlyne, D. E. Influence of complexity and novelty in visual figures

on orienting responses. Journal of Experimental Psychology,
1958(b), §§, 289-296.

Berlyne, D. E. Conflict, Arousala and Curiosity. New York: McGraw-
Hill, 1960.

Berlyne, D. E. Uncertainty and epistemic curiosity. British Journal
of Psychology, 1962, pg, 27-34.

Berlyne, D. E. Complexity and incongruity variables as determinants
of exploratory choice and evaluative ratings. Canadian Journal

of Psychology, 1963, 11‘3), 274-290.

Bruner, J. S., Olver, P., Greenfield, M. et al. Studies in Cognitive
Growth. New York: Wiley & Sons, Inc., 1966.

Brown, J. S. The Motivation of Behavior. New York: McCraw-Hill,
1961.

Burgess. cited in Berlyne, D. E. Conflict and information theory
variables as determinants of human perceptual curiosity. Journal

of Experimental Psychology, 1957,.23, 399-404.

Butler, R. A. Incentive conditions which influence visual explora-

tion. Journal of Experimental Psychology, 1954, 48, 19-23.

Cantor, G. N. Responses of infants and children to complex and novel
stimulation, in Advances in Child Developpent and Behavior I, L. P.
Lipsitt and C. C. Spiker (eds.), New York: Academic Press, 1963,
PP. 2'30.

55

56

Cantor, J. H. and Cantor, G. N. Observing behavior in children as
a sunction of stimulus novelty. Child Development, 1964, pp,
119-128.

Charlesworth, W. R. Cognition in infancy: where do we stand in the

mid-sixties? Merrill-Palmer Qparterly, 1968, 14, #1.

Clapp, W. F. and Eichorn, D. H. Some determinants of perceptual inves-

tigatory responses in children. Journal of Experimental Child
Psychology, 1965, ;, 371-388.

Piske, D. W. and Maddi, S. R. Functions of Varied Experience.
Homewood, 111.: Dorsey Press, 1961.

Gellerman, L. W. Chance orders of alternating stimuli in visual
discrimination experiments. Journal of Genetic Ps cholo , 1933,
43, 207-208.

Gibson, E. J. Perceptual development, in Child Psychology, H. W.
Stevenson (ed.) Chicago: University of Chicago Press, 1963,
PP e 144-195 9

Harris, L. Effects of relative novelty on children's choice behavior.

Journal of Experpppntal Child Psychology, 1965, ;, 297-305.

Barrie, L. Effects of relative novelty on preschool children's choice
of toys, choice latency, and visual orientation. 1967(a) Paper
read at 1967 meetings for the Society for Research in Child
Development, New York City, March, 1967.

Harris, L. Effects of amount of relative novelty on children's choice
behavior. Psychonomic Science, 1967(b), p, 319-320.

Harvey, O. J. Motivation and Social Interaction: Cognitive Determi-
nants. New York: Ronald Press 00., 1963.

Rays, W. L. Statistics for Psychologists. New York: Holt, Rinehart,
& Winston, Inc., 1963.

Hebb, D. 0. On the nature of fear. Psychological Review, 1946, pg,
259'276o

Hebb, D. 0. Organization of Behavior. New York: Wiley & Sons,
Inc 0 g 1949 0

Johnson, G. Effects of stimulus complexity and familiarization on
free-looking times of children and adults. MA thesis, 0. of
Minnesota, 1968.

Kagan, J., Beaker, 3., Hen-Tov, A., Levin, J., & Lewis, M; Infant's
differential reactions to familiar and distorted faces. Child
Developppnt, 1966,';1, 519-532.

57

Lintz, L. M., Starr, R. H., and Medinnus, G. R. Curiosity reward
in children. Psychological Reports, 1965, 16, 1222.

Lore, R. R. Some factors influencing the child's exploration of
visual stimuli. Unpublished doctoral dissertation (1965) U. of
Tennessee.

Maddi, S. R. Exploratory behavior and variation seeking in man in

Functions of Varied Experience. D. W. Fiske & S. R. Maddi (eds.)
1961.

McCall, R. and Regan, J. Attention in the infant: effects of complexity,
contour, perimeter, and familiarity. Child Developpent, 1967,.18,
939-952.

Maw, W. and Maw, E. Information recognition by children with high and
low curiosity. Educational Research Bulletin, 1961, 11, 197-201.

Maw, W. and Maw, E. Measurement of Curiosity in Elementary School
Children: A Review. Moravia, N.Y.: Chronical Guidance Publica-
tions, Inc., undated.

‘Miller, D. R. Motivation and Affect. in Handbook of Research Methods
in Child Developpent, P. Mussen (ed.), New York: Wiley & Sons, Inc.,

Munsinger, H. Tachistoscopic recognition of stimulus variability.

Journa1 of Experimental Child Psychology, 1965, 1, 186-191.

Munsinger, H. and Reason, W. Uncertainty, structure, and preference.
Psychological Monoggaphs, 1964, 18, no. 586.

Munsinger, H. and lessen, W. Stimulus variability and cognitive change.
Psychological Review, 1966, 13, 164-178.

Munsinger, 8., Reason, W., and lessen, M. Age and uncertainty:
developmental variation in preference for variability. Journal of

Expegimentel Child Psychology, 1964, 1, 1-15.

Piaget, J. The Origins of Intelligence in Children. New York: The
Norton Library, 1952.

Pielstock, N. L., Woodruff, A., and Bond. Exploratory behavior and
curiosity in two age and ability groups of children. Psychological
Reports, 1964, 13, 831-838.

Smock, C. D. and Holt, B. G. Children's reactions to novelty: an
experimental study of "curiosity motivation". Child Develo ment,
1962,111, 631-642.

Swan, C. Individual differences in the facial expressive behavior of
preschool children: a study by the time sampling method. Genetic

Psychology Monoggaphs, 1938, 19, 557-650.

58

Welker, W. I. An Analysis of Exploratory and Play Behavior in Animals,

in Functions of Varied Experience. D. W. Fiske & S. R. Maddi (eds.),
Homewood, 111.: Dorsey Press, 1961, pp. 175-226.

Welker, W. I. Effects of age and experience on play and exploration

of young chimpanzees. Journal of Qompapative and Physiological
Psychology, 1956(a), 42, 223-226.

Welker, W. I. Some determinants of play and exploration in chimpanzees.

Journal of Copparative and Physiological Psychology, 1956(b), 42,
84’89o

Welker, W. I. Variability of play and exploratory behavior in chim-
panzees. Journal of Com arative and Ph siolo ical Ps cholo ,
1956(c), 42, 181-185.

Appendices

59

Appendix A*
(1) Congruous Drawings

(2) Incongruous Drawings

 

*All drawings are reduced to approximately one-third or one-fourth the
original sise.

Congruous Drawings

 

 

 

 

 

 

 

 

 

 

 

APPENDIX B

ORDER OF THE PICTURES AMONG

THE FOUR BOOKS OF PICTURES

 

Original
Picture
Numbers Picture Book I Book II Book III Book IV
21 Elephant-
plane 1 24 20 20
_ 3 Horse 2 7 12 21
23 Squirrel-
elephant 3 11 23 22
29 Elephant-
turtle 4 23 19 16
5 Deer s 10 1 24
17 Turtle-
elephant 6 26 30 25
13 Jet 7 2 17 9
27 Car-
flamidgo 8 15 15 8
14 Rabbit 9 17 21 26
15 Fish 10 8 7 14
18 Boy-deer 11 27 5 7
l9 Horse-turtle 12 12 2 10
4 Turtle 13 29 27 3
2 Flamingo 14 5 11 13
25 Car-rabbit 15 21 25 19
16 Flamingo-
elephaht l6 6 9 15
3 Squirrel 17 14 18 27
26 Fox-duck 18 9 29 29
12 Giraffe 19 13 24 11
6 Elephant 20 20 22 4
21 Boy-giraffe 21 l 28 2
22 Bird-deer‘ 22 16 3 3O
1 Bird 23 25 6 23
10 Wolf 24 22 8 17
11 Duck 25 18 4 18
20 Horse-bird 26 4 16 28
9 Car 27 30 26 1
23 Elephant-
giraffe-bird 28 19 10 5
30 Dog-fish 29 3 13 12
7 Boy 30 28 14 6

APPENDIX C

LONGER MEAN DURATION OF LOOKING TIME PER TRIAL BLOCK

FOR EACH INDIVIDUAL SUBJECT

 

l

 

 

 

 

 

 

 

H
:3 hihlhat) (JCDPdF4P4CJFdhdF4P4Pi
.8
O
O
E:
H HHHH HUHHUHUHUHU
H H
m
H
m
E-0
H UHHH HUUHUHHUHHU
0
'U
m
to
I m HNMQ’ OHNMGM‘DI‘flOO
.: ml HHHHHHHHHHN
U
H
:3
0
he
H
s OHOO UUHHOHOHHHO
.8
0
O
3‘
H HHDH HHOHOHUUUHH
H H
G
44
H
E-I
H HHHH UUHOOHOHHUH
o
'0
m
u
N
I m Humx't Or-INMs'tn‘or‘wmo
g ml HHHHHHHHHHN
O
0
0
to
H
.3 u HUHO UHHHUHOUDHU
O
H
as
H
.3 : HOHH OHOHHUOHOUH
:4
E-I
H HUHH HHHOHOHHHOH
c:
0
u
u
2'.
H m '4ch c—cumonohoooso
0 col papapapapapd—apapapaca
'0
I:
”4
H

 

 

ins-La hi

IA...

AUG 1 q 7989

MICHIGAN STATE UNIV RSITY

Illl w

3 1293

ueanams
3

3046 680