u—v

 

1"“

3W 2W ‘- I ’
' '. .’ .3 "71h " I

' . |'d|(n\\r1”:’lw ’3": t1! ' U 4 $3.? 1“.“- 1|;é111}:’
.. I‘ I "

. é". I”! |‘1p};\l‘i::l:’.“‘ “1;“:

, .‘ ' 2,: g"; ;;.f"a’.'§. .“TI£.: .:

. ‘ 'I:I.itu’lg " l.’ ;|1.Tli'.':"'
.1. ,:;>,<I3§j:£.9i};-"'I' mm!

In“ If'fl" 'I . .
1" ‘ I‘O.I:I'I"‘i ‘ ‘ '

.' 913334:
"iII- ' 13411“,

i ' ......

.....

v - ‘
IIIIIIIIIII
_ I
,,,,,,

{W ”4‘;
III“

.32?!" {‘1‘ fif‘fif a I,” ‘ V
w 'fih‘t 3 £‘ . E‘:-‘4I
I Ii‘:‘( 1, Adi?
15"» ' , III“
”.311 I ”P I
I II‘ I f a?

JE.:I';I:L"5I-""I ‘v

«u I JWM

' I l " l1! :{3‘1’};
’ ?' 'Il

IE."
-_I

'1‘ P5; :" {”12" I ;; LV'I'Y :‘ 1.“ . I . . ‘. ':. 5'13 “3,. 14$:sz .L. :.
H ”b, L “I“?! .‘M rd?” 3:171. . fl: ”VII ‘3‘“ .
II "I"; ”g _',; '12:? ‘I .H‘ I“ A'W'I‘f: ‘23 .
f

. IinggiI

.35
”—3"?

Q.

4- .4%;+

'. . we"

 

'fiJ §%””@i'
III Ir. .‘ .7 ‘Z‘I'bl
*VHW’ .I‘I‘ II“: Ii” ’3' '; .' J”? .‘
(II. M ., "I
, ‘IW'I~
gig,” I! Ml H ’

*c—fi. a-
<
. "

 
   

“I! “i .I ,I 15”} 8W

14"“ ‘
'II WW?
' I JI‘I‘I'

' 4 L-._'..—.J~.

LIBRARY ‘1

Michigan State
University

 
     

This is to certify that the
thesis entitled
THE MEDIATING ROLE OF COGNITIVE

DEVELOPMENT IN ACHIEVEMENT
JUDGMENT AND BEHAVIOR

presented by

DeWayne Moore

has been accepted towards fulfillment
of the requirements for

LAW—degree inPSJChOIOEV '

«égM WZ,
Major 'professor

Date J119/77

0-7639

THE MEDIATING ROLE OF COGNITIVE
DEVELOPMENT IN ACHIEVEMENT

JUDGMENT AND BEHAVIOR

BY

DeWayne Moore

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

1977

ABSTRACT
THE MEDIATING ROLE OF COGNITIVE

DEVELOPMENT IN ACHIEVEMENT
JUDGMENT AND BEHAVIOR

BY

DeWayne Moore

Eighty children, 40 males and 40 females from the
first and third grades, participated in two 15 minute
sessions. In the first session, children evaluated the
achievement behavior of hypothetical others. In the
second session their level of cognitive maturity was
assessed and each child individually performed an embedded
figures task. Half the children were given information
linking their performance to the internal, stable factor
of ability. All children were told that they had not found
all the hidden figures (failure). Subsequently, their
affective and expectancy responses were assessed.

Although the attempt to manipulate attributions to
ability was not effective, an internal analysis revealed
that both expectancies and affective responses were
related to causal attributions. Level of cognitive devel-

Opment, included in a multiple correlation with causal

DeWayne Moore

attributions to predict expectancies and affect, did not
improve prediction.

Results of the achievement judgments revealed that
effort, ability and outcome were influential evaluative
determinants for both age groups and that even children at
the preoperational level used the three dimensions to

evaluate the performance of others.

To my mother and grandparents

ii

ACKNOWLEDGMENTS

The children, parents, teachers, and principal,
Mr. Canaday, of Elliott School, deserve my special thanks
for their assistance and cooperation.

I wish to express my appreciation to Dr. Barbara
Riemer, my committee chairperson, for her guidance and
well-timed support and to Dr. Ellen Strommen and Dr. Elaine
Donelson for their thoughtful and insightful advice.

In addition, Maureen Hart, Jane Dritsas, and Marc
Duffy served as testers and scorers; their assistance was

vital and is greatly appreciated.

iii

TABLE OF CONTENTS

LIST OF TABLES . . . . . . . . . . . . . . . . . . .

LIST OF FIGURES . . . . . . . . . . . . . . . . . .

LIST OF APPENDICES . . . . . . . . . . . . . . . . .

CHAPTER

I 0 INTRODUCTION 0 I O O O O O O O O O O O O O 0

An Attributional Theory of Achievement .
Sex Difference in Expectancies . . . . .
Cognitive Development . . . . . . . . . .
Development of the Attributional Process
Summary of Hypotheses . . . . . . . . . .

II. METHOD . . . . . . . . . . . . . . . . . . .

Subjects . . . . . . . . . . . . . . . .
Instruments . . . . . . . . . . . . . . .
Procedure . . . . . . . . . . . . . . . .
Cognitive Measures . . . . . . . . . . .
Scoring of Cognitive Measures . . . . . .
Achievement Task . . . . . . . . . . . .
Apparatus and Materials . . . . . . . . .

III. RESULTS . . . . . . . . . . . . . . . . . . .

Achievement Judgment Stories . . . . . .
Self Evaluations . . . . . . . . . . . .
Cognitive DevelOpment . . . . . . . . . .

IV. DISCUSSION . . . . . . . . . . . . . . . . .

APPENDICES

conCIuSion O O O O O O O O O O O O O O 0

LIST OF REFERENCES . . . . . . . . . . . . . . . . .

iv

Page

Vi

vii

i—J

67

7O

77

Table

l.

11.

12.

LIST OF TABLES

Means and Standard Deviations of Achieve-
ment Judgment Scores . . . . . . . . . .

Summary of the 2 (Sex of Subject) x 2
(Grade Level) x 2 (Effort) x 2 (Ability)
x 2 (Outcome) Repeated Measures ANOVA .

Means and Standard Deviations on Self-
Evaluation Measures . . . . . . . . . .

Summary of 2 (Grade) x 2 (Sex) x 2 (Sex
of E) ANOVA Performed on the Initial
Expectancies . . . . . . . . . . . . .

Summary of 2 (Sex) x 2 (Grade Level) x 2
(Attribution Instructions) ANOVA Per-
formed on Attributions to Ability . . .

Means and Standard Deviations of Ability
Attributions by Grade and Sex . . . . .

Correlations of Attributions with
Expectancy DrOps . . . . . . . . . . . .

Summary of the 2 (Sex) x 2 (Grade Level)
ANOVA Performed on Expectancy Drops . .

Correlations for Attributions with Affect

Summary of One-Way (Grade Level) ANOVA
Performed on the Affect Ratings . . . .

Mean Resultant Effort Score for Subjects
at Different Cognitive Levels . . . . .

Zero-Order Correlations Used in Multiple
Correlations for Hypotheses 12 and 13 .

Page

37

38

42

43

46

46

47

48
50

51

53

54

LIST OF FIGURES
Figure Page
1. Resultant Rewards Allocated for Effort,
Ability, and Outcome as a Function of

Grade Level . . . . . . . . . . . . . . . . . 40

2. Initial Expectancies of Males and Females
as a Function of Grade Level . . . . . . . . . 44

vi

LIST OF APPENDICES

APPENDIX Page
A. ACHIEVEMENT JUDGMENT STORIES . . . . . . . . . 70
B. INSTRUCTIONS FOR HIDDEN FIGURES TASK . . . . . 73
c. DEPENDENT MEASURES FOR HIDDEN FIGURES TASK . . 74

vii

CHAPTER I

INTRODUCTION

Cognitive maturity is known to have important
effects on social behavior (Lee, 1971; Tomlinson-Keasey &
Seasey, 1974; Selman, 1971; Feffer & Gourevitch, 1960).
This research has been guided by the assumption that cog-
nitive change in itself will be one critical determinant of
how children respond to their social environment. Only
recently, however, have achievement-related behaviors begun
to be studied within this framework (Weiner & Peter, 1973;
Salili, Maehr, & Gillmore, 1976).

The following literature review is designed to
examine the cognitive develOpmental approach to achieve-
ment. The attributional analyses of achievement proposed
by Weiner and his associates (Weiner, Frieze, Kukla, Reed,
Rest, & Rosenbaum, 1971) and Piaget's theory of cognitive
develOpment are reviewed. Finally, the research attempting
to integrate cognitive developmental theory with
attribution theory is critically examined, and the neces-

sity of further research is discussed.

An Attributional Theory of Achievement

Adopting Heider's (1958) notions, Weiner and his
associates (Frieze & Weiner, 1971; Weiner, Frieze, Kukla,
Reed & Rosenbaum, 1971; Weiner & Kukla, 1970) have devel-
Oped an attributional model of achievement motivation. The
model postulates four causal attributions that "mediate
between antecedent stimulus—organism transactions and
ensuing achievement behavior" (Weiner et al., 1971, p. 96).
The four causal elements are ability, effort, task diffi-
culty and luck. The theory postulates that in attempting
to explain the prior outcome (success or failure) of an
achievement related event, "the individual assesses his
own or the performer's ability level, the amount of effort
expended, the difficulty of the task, and the magnitude
and direction of experienced luck" (Weiner et al., 1971,
p. 96). Much of the research generated by Weiner's model
has focused on the relationship between causal attributions
and two achievement-related behaviors: expectations of
success and affective responses to achievement outcomes.

The first research to examine the relationship
between causal attribution and expectations of success was
conducted by Rotter and his associates (Rotter, 1966;
Phares, 1957; Lefcourt, 1966). Rotter (1966) cited a
number of studies supporting his hypotheses that changes
in expectancy are a function of whether the person perceives

internal or external control over the Situation. Lefcourt

(1966) defined the control poles: "Internal control refers
to the perception of positive and/or negative events as
being a consequence of one's own actions . . .; external
control refers to . . . events as being unrelated to one's
own behavior . . . and therefore beyond personal control"
(p. 207). Rotter cited a study by Phares (1957) which is
representative. In this experiment, subjects were given a
discrimination task that was ambiguous with respect to the
Objective determinants of success and failure. The outcome
in one condition was described as determined by skill
(internal control), while in a second condition performance
was said to be entirely a matter of luck (external control).
Expectancy of success was inferred from the number of chips
subjects were willing to bet on their next performance.
Phares found that there were typically more expectancy
shifts, and the shifts were of greater magnitude in the
skill condition than in the chance condition.

Weiner et a1. (1971), subsequently reformulated
Rotter's hypotheses. They noted the causal determinants
of outcome listed by Rotter (1966) differed not only along
the dimension of locus of control (internal vs. external),
but also along a stability dimension (fixed vs. variable).
According to this model, the results of an achievement
situation can be attributed either to internal or external
factors (locus of control) and to fixed or variable

factors (stability). Estimates of ability and task

difficulty are relatively fixed, while estimates of
intended effort and anticipated luck are free to vary from
trial to trial. In addition, effort and ability are prOp-
erties internal to the person while task difficulty and
luck are external factors.

Several studies have supported Weiner's model that
stability and not locus of control is the determining
dimension (Weiner, Nierenberg, & Goldstein, 1976). Meyer
(1970; reported in Weiner, 1972) was the first to compare
expectancy shifts between stable and unstable attributions.
Male high school students were given five repeated failures
at a digit-symbol substitution task. Following each trial
subjects attributed their failure to low ability, bad luck,
low effort, or task difficulty. Next, they estimated their
probability of successfully completing the next trial.
Expectancies decreased most following failure when
attributions were to low ability and/or task difficulty
than when attributions were to lack of effort or bad luck.

A study by McMahan (1973) essentially replicated
the results using a different experimental procedure. He
found, using a correlational design, a positive relation-
ship between the stable attributes of ability and task
difficulty and subsequent expectancy following success and
a negative relationship following failure, while the
relationships between effort and luck attributions and sub-

sequent expectancy tended to be negative following success

and positive following failure. As McMahan noted, the
obtained correlations between effort and task attributions
and subsequent expectancy were in the opposite direction
from that predicted by the one-dimensional analysis of
Rotter. Together, these results suggest that the stability

dimension is the more salient in an achievement context.

Sex Difference in Expectancies

Several studies indicate that females have lower
expectancies for success than males. Crandall (1969)
demonstrated in a series of studies on various age groups
and on a variety of tasks the generally low expectancies
of females, compared to males. Her samples included ele-
mentary school children who gave expectancy estimates for
their performance at novel intellectual tasks; eighth
graders who were asked to state how well they expected to
do at a digit-symbol matching task; college-aged people
from the Fels longitudinal sample who guessed their per-
formance at a geometric task. The results were consistent:
females had lower expectations than males in all situ-
ations.

Other researchers have replicated Crandall's find-
ings with a variety of age groups and tasks. Using 10
year old subjects, Montanelli and Hill (1969) found that
boys expected to do better than girls at marble dropping
games. Parsons (1974) found similar differences in

expectancies in a group of children 6 1/2-8 years old.

The task consisted of identifying objects hidden in a
series of pictures. Furthermore, high school boys antici-
pated more favorable performances than their female class-
mates on verbal intelligence tests (Brim, Goslin, Glass,

& Goldberg, 1969), and college males had higher expec—
tations for success at anagrams (Feather, 1969). Finally,
Nichols (1975) found that fourth grade boys had higher
expectancies than girls at an angle matching task.

Based on the Weiner et a1. (1971) model, researchers
have put forth two related hypotheses to account for sex
differences in expectancies. One line of research has
attempted to account for differences in expectancy in
terms of the different attributional tendencies of males
and females. For example, Dweck and her associates (Dweck &
Reppucce, 1973; Dweck & Bush, 1976) have found that girls
are relatively more likely to attribute failure to the
stable factor of lack of ability. Boys, on the other hand,
are more likely than girls to attribute failure to unstable
factors, such as lack of effort (Dweck & Reppucci;

Dweck & Bush) or bad luck (Nichols, 1975). As discussed
above, attributions following failure to stable factors
such as ability and/or task difficulty are associated with
larger drops in expectancy than are attributions to
unstable factors such as effort and/or luck. Indeed,
Nichols found that females, who had lower expectancies than

males, also differed from males in their attributional

patterns. After a practice session in which half the sub-
jects failed and half succeeded, boys reported higher
expectancies than girls. Attributions were also measured.
Nichols found that girls, but not boys, attributed failure
to poor ability more than success to good ability. It was
also found that boys more than girls attributed failure to
bad luck; and boys, but not girls, had higher luck
attributions after failure than after success. Thus, the
sex differences in expectancies were paralleled by sex
differences in attributions.

A second but related explanation of sex differences
in expectancies rests on the assumption that the stability
of causal attributions made about a performance is related
to whether or not the outcome confirms or disconfirms prior
expectations. The lower expectancies of females may result
in success being an unexpected outcome and failure being an
expected outcome, while for males, success may be an
expected outcome and failure an unexpected outcome.

Feather (1969) has hypothesized that unexpected outcomes

are attributed to variable factors and expected outcomes

to stable internal factors. More specifically, Feather
found that subjects with high expectations of success tended
to attribute success to ability (stable attribute) and
failure to bad luck (variable attribute); on the other

hand, subjects with low expectations of success tended to

attribute success to good luck and failure to lack of

ability. Subsequent studies by Feather and Simon (Feather
& Simon, 1971a, 1971b, 1972; Simon & Feather, 1973) have
supported this notion.

The stability of attributions made about one's
performance is thus partially a function of the difference
between the actual outcome and the initial expectancies.
Valle and Frieze (1976) have develOped a mathematical model
relating initial expectation and causal attributions.
Unexpected outcomes are attributed to unstable causes and,
therefore, have less weight in determining future pre-
dictions; expected outcomes are attributed more to stable
causes and tend to support and reinforce original expec-
tations. When persons with low expectancies fail on a
task, an expected outcome, they tend to attribute it to
lack of ability. This cycle maximizes the negative effects
of failure and minimizes the positive effects of success.
Low expectations, then, are self-perpetuating since they
lead to attributions which maintain their accuracy.

To summarize, sex differences in initial expec-
tancies of success lead to sex differences in attributional
patterns which result in differential expectancy drOps.
This study further examined this cycle.

Weiner et a1. (1971) have also hypothesized that
while the stability of the attributions following success
or failure influence the individual's expectancy of future

success and failure, the locus of control influences

affective responses to success and failure. For example,
attributions to internal factors (ability or effort) are

assumed to result in greater positive or negative affect

than attributions to external factors (task difficulty or
luck).

Weiner (1974) reviewed several studies indicating
that locus of control influences affective responses to
success and failure in adults. In addition, Weiner and
Kukla (1970) found that affective reSponses to feedback
were a function of the degree to which the outcome was
seen as internally determined with attribution to effort
being particularly important. Nichols (1975), however,
found that pleasure with success correlated with attribution
of success to high ability, but not with attribution to
high effort. This study provided further evidence on the
relations between attributions of ability and effort and

affective reSponses.

Cognitive Development
Since researchers examining the development of the
attribution process have assumed that concrete Operations
are central for the emergence of the information-
attribution-affect and expectancy link (Weiner, 1974;
Weiner & Peter, 1973; Ruble et al., 1976), cognitive-
developmental theory is reviewed, with particular emphasis

on the concrete Operational period.

10

The cognitive-developmental approach, exemplified
in the writings of Piaget (1932, 1926, 1928, 1929), repre-
sents a major theoretical framework for conceptualizing
change processes. Piaget views develOpment as forming an
invariant sequence with each developmental stage repre-
senting a qualitative different mode of thought. Because
earlier stages are necessary parts of their successors, the
individual must pass through all preceding stages before
he can move on to the next one (Flavell, 1963).

Piaget prOposed several important changes that
accompany the acquisition of concrete Operations. One
particularly important achievement is the ability to con-
serve. Conservation of mass and conservation of liquid,
for example, have dramatically illustrated children's
emergence from the preOperational to the concrete Opera-
tional level where reversible thought Operations are
exhibited by the ability to conserve. One operating prin-
ciple here is the ability to decenter, i.e., "to take into
account features which could balance or compensate for the
distorting and biasing effects of a single centration
(Flavell, p. 157). As Flavell points out, the tendency
to center, i.e., "focus attention on a single, striking
feature of the object of its reasoning to the neglect of
other important aspects, and by so doing to distort the
reasoning" (p. 157), is one of the most salient character-

istics Of preoperational thought.

11

Another important series of changes takes place in
children's classification of Objects. Piaget (1952)
investigated children's grasp of the relations between the
whole and the parts of a class. Among this group of
Operations, Piaget views the "relationship of inclusion"
as being the principle stumbling block for the cognitively
immature child. Piaget (1952) has stated that the pre-
Operational child "cannot establish a permanent inclusion
between the whole and the parts: as soon as the whole is
divided, even in thought, the parts cease to be included
in it and are merely juxtaposed without synthesis" (p. 171).

Finally, with the acquisition of concrete operations,
children begin to overcome the egocentrism of the pre-
Operational period. Preoperational children repeatedly
demonstrate an inability to take a perspective different
from their own. This form of egocentrism results from a
tendency to confuse thinking about an event and the event
itself. PreOperational children tend to assimilate reality

to their own perspective (Flavell, 1963).

DevelOpment of the Attributional Process
Weiner and his associates (Weiner, 1974; Weiner &
Peter, 1973) have attempted to integrate cognitive-
developmental theory with attribution theory. They sug-
gest that certain universal developmental patterns may
underlie achievement cognitions and achievement behaviors.

As Weiner and Peter (1973) state, "Analysis of the cognitive

12

components in the achievement system supports the belief
that a cognitive-developmental approach is applicable in
the achievement domain. Individuals concerned with
achievement-related goals have forward-looking time per—
Spectives, realistic yet positive expectations of success,
and a unique pattern of causal ascriptionsfor success and
failure. Thus achievement motivation is associated with a
particular pattern of cognitive functioning" (p. 291).
Weiner (1974) further notes that the attributional
model of motivation is "likely to be affected by the
cognitive maturity of the actor" (p. 43). Thus cognitive
change in itself may be a critical determinant of the
child's achievement-related behaviors. One of the cognitive
characteristics of the preOperational child is an inability
to separate the physical world from his own motivation.
This, along with one of the most pronounced characteristics
of preoperational thought, the tendency to "center"
(Flavell, 1963) suggests as Parsons, Ruble, Hodges, and
Small (1976) pointed out, that the young child may not use
past experience in predicting future events. According to
Inhelder and Piaget (1958), it is not until the stage of
concrete operations that the child becomes capable of
relating temporarily separated events to each other and
avoids focusing on single events. This inability to
decenter would make it difficult for the preOperational

child to use past experience in predicting future outcomes.

13

One implication of a cognitive developmental analy-
sis is that if children do not use serial information to
form expectancies until they enter the concrete Opera-
tional period, then sex differences in expectancies would
not be expected until children acquire concrete Operations.
And indeed, research by Parsons and her associates (Parsons,
1974; Parsons and Ruble, 1975; reported in Parsons, Ruble,
Hodges, and Small, 1976) provides indirect support.
Unfortunately, age rather than cognitive development was
relied on as the independent variable. Their research, in
any event, suggests that sex differences in expectancies
for success do not develOp until sometime after the pre-
school years.

An additional implication of cognitive immaturity
results from the inability to use serial information noted
above and the tendency to confuse the psychological and
physical realms (Flavell, 1963). Young children desiring
to succeed may assume that the wish to succeed and actually
succeeding are the same. Consequently, young, preopera—
tional children would be expected to have higher expec-
tancies for success than older children who have mastered
concrete Operations. Again, indirect support for this
prediction comes from the work of Parsons and Ruble
(Parsons, 1974; Parsons & Ruble, 1972). They found for
subjects ranging in age from 3 1/2 to 10 that expectancies

for success decreased as a function of age.

14

Finally, if as Flavell (1963) notes, "the preopera-
tional child is confined to the surface of the phenomena
he tries to think about, assimilating only those super-
ficial features which clamor loudest for his attention"

(p. 157), the young child would be expected to focus on
concrete and external aspects of achievement situations.
Consequently, the outcome of an event is likely to be the
salient feature. Indeed, in the area of moral development,
this cognitive immaturity is reflected in the child's
resolution of moral dilemmas. Typically, he focuses on
the physical aspects of a moral dilemma, such as whether
it was a big or little lie, or how much physical damage was
done (Piaget, 1932). Likewise, an achievement event may
be considered simply in terms of success or failure. With
increased cognitive maturity, however, there should be an
increased capacity to consider the event in a more complex
fashion.

Such an analysis has received some support (Ruble,
Parsons, & Ross, 1976; Weiner & Peter, 1973; Salili, Maehr,
& Gillmore, 1976). Ruble et a1. (1976) hypothesized that
young children would not use locus of control and outcome
information in ways suggested by research with adults. In
their study, children ranging in age from 6 to 10 were
asked to perform a task and were given patterns of infor-
mation which were expected to produce internal or external

attributions. The children's self-evaluative reactions

15

about their performance were then measured. The results
indicate that the younger children's affective responses
were not affected by the information they were given.
The affective responses Of the six year old children
depended directly upon the outcome of their behavior. The
affective responses Of the Older children—-8 to 10 years
of age--, however, were greater for those children who
received information indicating internal attributions.
One explanation for these results may be the young, pre-
Operational child's tendency to center. The younger,
cognitively immature children focused on the outcome of
the event to the exclusion of causal attributions. Such
an explanation, while consistent with the results, seems
premature, however, since cognitive develOpment was not
measured.

Weiner and Peter (1973), likewise, found age-
related differences in the influence of causal attributions
on achievement judgments. Weiner and Peter asked children
4-18 years of age to make judgments about achievement-
related outcomes using three evaluative dimensions:
effort, ability and outcome. Subjects were told that
children in a classroom had been assigned a puzzle to
complete. The pupils were described as high or low in
ability, high or low in effort expenditure, and as succeed-
ing or failing at the task. The subjects were given all

eight possible combinations Of information. They were

16

then asked to evaluate the pupils. The evaluation could
range from +5 (reward) to -5 (punishment). First, it was
found that achievement judgments became more positive as
age increased. Second, among the younger children, 4-6
years of age, achievement evaluation was primarily deter-
mined by the outcome Of an action. Success was rewarded,
while failure was punished. Then, with increasing age,
the amount of effort expended became an influential deter-
minant of evaluation. In a review of this experiment,
Weiner (1974) thus concluded, "the consequences of causal
ascriptions to effort change with cognitive development"
(p. 45).

But here again, as in the Parsons and Ruble studies
(Parsons, 1974; Parsons & Ruble, 1972; Ruble et al., 1976;
Parsons & Ruble, 1975; reported in Parsons et al., 1976),
the children's level of cognitive development was not
assessed. Age, rather than cognitive level, was relied on
as an independent variable. Because age was thus equated
with cognitive level in these studies, conclusions which
cite the centrality of cognitive development seem pre-
mature. The demonstration Of a central role for cognition
requires that the stage of cognitive development and not
age be the major variable.

The present research was designed to examine the
questions raised by the preceding discussion. First,

since attributions mediate expectations and affect, both

17

can be changed by manipulating the perceived causes of a
performance. Indeed, Dweck (1975) has demonstrated the
efficacy of manipulating causal attributions, though in a
slightly different context. She attempted to teach children,
who had previously been diagnosed as "helpless," to take
responsibility for failure and attribute it to lack of
effort. The hypothesis was that this would lead to
increased persistence on the part of these children when
compared with another group of helpless children who
received only success experiences during the training
period. This group did not receive the attribution train-
ing. The results of the investigation supported the hypoth-
esis. The children who were taught to attribute failure
during training to insufficient effort showed a significant
increase in persistence from pretraining to posttraining
when compared to subjects who received the success only
treatment. This demonstrated the mediating effect of
causal attributions on persistence.

In a study by Riemer (1975), all four causal ele-
ments were manipulated as independent variables in order
to determine the relationship between locus of control and
affect, and stability and expectancy. The results indicate
that locus of control influenced the subjects' affective
responses. The affective responses were higher for the
internal attributes Of effort and ability than for the

external attributes of task difficulty and luck. The

18

stability dimension, however, did not have the predicted
affect on expectancy of success. As a possible explanation
for these results, it should be noted that this study dealt
only with attributions for success. A number of studies
discussed above (Dweck & Bush, 1976; Dweck & Repucci, 1973;
.Parsons, 1974; Nichols, 1975; Meyer, 1970) suggest that
causal attributions following failure may be a more influ-
ential mediator Of achievement-related responses. More
specifically, Nichols found differential attributional
patterns for high expectancy-males and low expectancy
females only after failure. In this study attributional
patterns following success could not account for the sex
differences in expectancies. Furthermore, Parsons found
that among 6 1/2-8 year olds, girls and boys responded to
success in a similar manner. In both cases, their expec—
tancies increased. However, they responded to failure
differentially. The girls drOpped their expectancies
more. Failure feedback may, therefore, be more salient in
the hypothesized stability and expectancy relationship.
The failure in the Riemer study to verify the predicted
relationship between the stability of the causal attribution
and the expectancy response could be due to manipulating
attributions for success only.

In order to examine the information-attribution-
expectancy and affect link, the present experiment attempted

to manipulate attributions to ability, a stable, internal

19

attribute. Furthermore, since attributions following
failure may be more influential in the information-
attribution-expectancy and affect link, all children
received failure feedback on a hidden figures task.

In addition, the Weiner and Peter (1973) study,
which asked children to evaluate the performance of others,
was replicated in order to determine the relationship
between the mediating affect causal attributions on both
achievement judgments and achievement behaviors. Finally,
the present study examined the influence of cognitive
development on the mediating influence of causal attri-
butions on expectancies and affect and achievement judg-

ments. Age was thus not relied on as a variable.

Summary of Hypotheses
In line with the preceding discussion the following

predictions were made.

Achievementgjudgments. The first three hypotheses

 

are based on results from Weiner and Peter (1973).

1. When judging the performance Of others, evaluation
will be primarily determined by the outcome of the
action. This will be true of both age groups.

2. However, among the older children the amount of
effort expended will be an influential evaluative

determinant.

20

3. Finally, achievement judgments will become more

positive with increasing age.

Self evaluations expectancy. Hypotheses 4 and 5

 

are based on the results from Parsons (1974).

4. Expectancies for success will be age-related. The
younger children will have higher expectancies
than the Older children.

5. Sex differences in expectancies will be age-related.
Specifically, the predicted higher expectancies
for success of males will be found only for the
older children.
Hypotheses 6a through 7b reflect the assumed role

of cognitive development.

6a. It is predicted that expectancies will drop follow-
ing failure. However, the expectancy drops of the
younger children will not be mediated by causal
attributions. Specifically, the younger children's
expectancies should not be affected by the instruc-
tions suggesting attributions to ability.

6b. Secondly, the expectancy drOps of the younger
children will not be related to the attributions
given following failure.

6c. Finally, the expectancies of the younger children
will drop more than the expectancies of the Older

children.

7a.

7b.

21

Among the Older children causal attributions will
be an influential determinant of expectancies.
Specifically, children who are led to believe that
failure is primarily determined by ability (stable
factor) should decrease expectancies of success to
a greater extent than children who do not receive
the attribution orienting instructions.

In addition, the expectancy drops of the Older
children will be related to the attributions given
following failure. The expectancy drops will be
positively related to the stable attributions and
negatively related to the unstable attributes.

Hypotheses 7c and 7d follow from the hypothesized

higher initial expectancies of third-grade males

(Hypothesis 5).

7c.

7d.

The expectancy drops of third grade males and
females will differ. Third grade females will
drOp their expectancies more than will third grade
males.

Third grade males and females will differ in their
attributions. Third grade females will give more
attribution to ability than will third grade males.

Hypotheses 8a through 9b reflect the assumed role

of cognitive develOpment.

22

8a. The affective responses of the younger children
will not be affected by the attribution orienting
instructions.

8b. The affective responses Of the younger children
will not be related to the attributions following
failure.

9a. The affective responses of the Older children
will be influenced by the attribution orienting
instructions. Greater negative affect will be
expressed by the older children in the ability
condition.

9b. The affective responses of the older children will
be related to the attributions made following
failure. The expectancies of negative affect will
be positively related to the internal attributions
and negatively related to the external attribu-
tions made following failure.

10. Finally, Parson (1972) and Ruble et al. (1976)
found an age-related increase in the impact of
failure information on affect ratings. Based on
these findings, it is predicted that the Older
children will be more negative about their affect

ratings after failure than the younger children.

Cognitive development. Finally, it was hypothesized

 

that the impact of concrete Operational thinking will be

reflected by qualitative changes in the mediating affect

23

of causal attributions on achievement judgment and achieve-
ment behavior (affect and expectancy). However, once the
impact of a major cognitive reorganization has been reflected
by changes in the mediating role of causal attributions, it
is not immediately obvious how cognitive develOpment would
continue to exert its influence. Consequently, there
should be an orderly fluctuation in the magnitude of the
relationship between cognitive develOpment and information-
attribution-achievement behavior link. The magnitude of
the relationship should be higher following major cognitive
reorganizations and lower during the intervals between
such cognitive reorganizations. Therefore, it was pre-
dicted that the correlation between cognitive develOpment
and the mediating affect of causal attributions on both
achievement judgments and behaviors would be greater among
children evidencing an initial mastery Of concrete opera-
tional thought than among those firmly established in con-
crete Operations. The following four hypotheses, therefore,
applied only to the younger children.
11. The mediating influence of effort on achievement
judgments will be related to the children's level
Of cognitive development.
12. The influence of ability attributions on expec-
tancies will be related to the children's cognitive

level.

13.

14.

24

Cognitive development will also be related to the
influence of ability attributions on affect.
Finally, high expectancies of success will be
negatively related to the level of cognitive

development.

CHAPTER II

METHOD

Subjects

Eighty children, 40 males and 40 females, were
selected from the first and third grade classes of a lower-
middle to middle-class suburb of Lansing, Michigan. Per-
mission was obtained from the school administrators to test
three classrooms at each grade level. Prior to the study,
parents were notified of their child's possible partici-
pation and permission slips were returned by the children.
The average age of the children was 6.72 years for the
first graders and 8.76 years for the third graders.

The teachers at each grade level served to briefly
introduce the experiment to the children, indicating that
the experimenters were from Michigan State University and
desired help with a project involving elementary students.
All children were asked whether they wanted to participate
and were given ample Opportunity to refuse. All children

agreed to participate.

25

26

Instruments

Achievement judgments. The achievement-related

 

judgments were adapted from Weiner and Peter (1973). The
situations involved a child working at a puzzle. The child
was characterized according to ability (present or absent),
"trying" (yes or no), and the consequences of the action
(completion and success or incompletion and failure).
More specifically, the subjects were told the following:
I am going to play a game with you, and in this
game I am going to tell you about some children.
After I tell you about each boy or girl, I want you
to put a check mark above either gold or red stars,
whichever you think the boy or girl should get. Gold
stars mean that you are pleased with the boy or girl
and that you want to reward him or her. Red stars
mean that you are not pleased with the boy or girl
and you want to punish him or her. You can check
either 1, 2, 3, 4, or 5 gold or red stars, depending
on how many you think the child deserves. Five gold
stars would be a big reward and 1 gold star would be
a little reward, while 5 red stars would be a big
punishment and 1 red star a small punishment. Do you
understand all that?
Now let's practice a little. Bernie is a little
boy who helped his mother with the dishes. What color

star would you give him? How many stars? Susie is a

27

little girl who would not clean up her room. What
color star would you give her? How many stars? Now

I am going to tell you about some other children.
These children are in school and the teacher has given
them a picture puzzle to put together. This is the
kind of puzzle in which you fit the pieces together.
Each child is supposed to put his or her puzzle
together before the bell rings. I'll tell you about
each child and you give him or her either gold or red
stars.

For all subjects, the instruction and the experi-
mental conditions were read aloud. The eight experimental
conditions judged are presented in Appendix A.

Each subject received a booklet containing the
eight stories (see Appendix A). Under each story were two
rows of numbers l-S. Directly below each row of numbers,
the apprOpriate number of stars were reproduced. The first
row contained gold stars, the second, red stars. Subjects
were asked to put a check mark above the color and number

of stars they thought the child in the story deserved.

Selection of cognitive measures. The tasks used

 

to measure cognitive development were selected to define
the different stages of development and to stay close to
Piaget's original concepts and data-gathering procedures
as possible. Since the subjects used were children 6-7

and 8-9 years of age, their levels of cognitive functioning

28

was assumed to range from the preOperational subperio-
through the concrete Operational subperiod. Since the
ability to perform reversible Operations on classes and
relations are essential processes in the emergence of
concrete Operational thought (Piaget, 1952, Ch. 2), two
tasks were selected as being representative of operations
performed upon logical classes and two tasks were selected
involving Operations performed upon the relations which may
exist between two classes.

The tasks selected were conservation of mass and
conservation of liquid, and two tasks dealing with the
child's ability to additively compose classes--the "wooden

bead" problem and the "children" problem.

Procedure

Each child participated in two 15 minute sessions.
The first session, during which the achievement judgment
stories were completed, consisted of groups of five same
sex and same age children randomly drawn from within a
classroom. The classroom from which the children were
drawn was randomly selected from among the total of six.
There were three classrooms for each grade. A few excep-
tions to this procedure occurred in order to avoid incon-
veniencing a teacher. This first session was conducted by
an experimenter of the same sex as the children.

After all subjects had completed this session, the

second session began. This session was conducted by a

29

male or female experimenter (counterbalanced for the sex
of subject). There were two male and two female experi-
menters. Four measures of cognitive development were com-
pleted first. Then the child was asked to do the achieve-
ment task. Each subject was tested individually, seated

across a table from the experimenter.

Cognitive Measures
All subjects were assigned the four Piagetian

tasks in the following order.

Conservation Of mass. Two play-doh balls identical

 

in size, shape and weight were placed on the table in front
of the child.

"DO the balls have the same amount of play-doh,
is there as much in this ball as in this one?" (As an
additional check, the following instructions were included.)
"If this was candy, and you eat this ball and I eat this
one, will we both have just as much to eat?"

The child was encouraged to make them the same if
he doubted the equality Of the balls. When the child
agreed that the two balls were equal, the experimenter
proceeded.

"If I made my ball of candy into a hot dog, will
I have just as much candy to eat as you have? Will we

both have the same amount of candy to eat?"

30

The experimenter actually rolled one Of the balls
into a hot dog while the child watched.

"Do we both have the same amount of candy to eat
now? Why?"

If the child had any questions or did not under-
stand the instructions, the experimenter repeated the

instructions.

Conservation of liquid. Two jars (A & A') of equal

 

size were placed before the child and water was poured
into each while the child watched. "If you were to drink
this water and I drink this, will we both have just as much
to drink?" If the child doubted the equality of the two
jars, water was poured into each until the child was
satisfied that both jars contained the same quantity of
water.

"Now, let's pour this (A') water into this one
(cylindrical jar). DO we both have the same amount of
water to drink? Do you have just as much to drink as I

have? Why?"

Children problem. The child was asked the follow—

 

ing questions: (a) How many children are there in your
class? (b) How many boys in the class? (c) How many girls
in the class? (d) Are there more girls (or boys depending
upon the sex of the child being questioned) or more children

in your class? Those children who answered incorrectly

31

were asked (e) What are children? and then question (d)
was asked again. The questions were repeated as often as

necessary to insure the child's attention and understanding.

Bead problem. The child was given a box containing

 

20 blue beads and 4 red beads, all of which were made of
wood. After ascertaining that the child was aware that
all the beads were made of wood, he was asked, "Are there
more wooden beads or more blue beads?" Those children who
answered incorrectly were asked, "Are all of the beads
made of wood?" And then the previous question was asked
again. The questions were repeated as Often as necessary

to insure the child's understanding and attention.

Scoring of Cognitive Measures

The concrete Operations task were tape recorded
and scored independently by two peOple according to the
theoretical criteria established by Piaget (1952) and used
by Elkind (1961a, b). For the children and bead problem,
the responses were categorized as follows: All responses
indicating the partial class was greater than the total
class were categorized as Stage I responses and assigned
a score of zero. Responses indicating that the partial
class was identical with the total class were categorized
as Stage II responses and assigned a score of 1. Those
answers which were "correct" were categorized as Stage III

responses and given a score of 2.

32

For the conservation tasks, each conservation
response, including both judgments and explanations, were
scored 2 and all nonconservation responses were scored
zero. A correct judgment without a reasonable explanation
was scored 1. The total scores for each child were summed,
yielding a composite index of scores ranging from zero to

eight. The inter-rater reliability was .98.

Achievement Task

A task fulfilling several criteria was needed for
the study. The task had to be sufficiently ambiguous
regarding the determinants of performance to yield an
effective manipulation of causal attributions. Further,
the outcome had to be under experimental control. Finally,
a task having some interest for the participants was
desired. With this in mind, subjects were asked to find
the hidden Objects in each of two hidden object pictures.
The subjects were divided into one Of two experimental
conditions: a control group and an ability instruction
group. Prior to attempting the task, subjects were read
one of two sets of instructions. The ability instruction
group were told that success or failure at finding the
hidden Objects was due to ability. The intent of these
instructions was to induce attributions to ability. The
instructions read to the control group were identical
except that no information about the causes of success or

failure was included (see Appendix B).

33

After the instructions were read, subjects were
asked how good they thought they would be at finding the
hidden animals. Subjects then began the task and were told
at the end of a designated time that they did not (failure)
find all the hidden animals in the picture. The degree of
difficulty of the pictures was such that all children found
three or four animals, at which point the experimenter
called "time." Time was called on the average after about
30 seconds for the first graders and after about 20 seconds
for the third grade children.

Following this trial, a questionnaire containing
the dependent measures was administered. The questionnaire
contained one expectancy measure, two affect measures, and
one question included to assess attribution. Data col—
lection was complete at this point. An additional trial
was given on which all children "succeeded" and were told

that they had done very well.

Apparatus and Materials

Dependent measure. The expectancy and affect

 

measures were adapted from the procedure used by Hill and
Dusek (1969). The child was presented with a bar graph
with a stick figure under each of the seven bars. The
bars decreased linearly across the page from left to right.
The subject was told that the child represented by the

stick and bar to the far left "does best at this task," and

34

the one to the far right "does poorest at this task." The
subject was then asked to point to the child he thought
he would turn out to be. The same principle was used for
the affective measures (see Appendix C).

The attribution measure was adopted from McMahan
(1974). Subjects attributed causality for their failure
by means of a set of six paired comparison questions, all
possible pairings of the four attributional factors. The
use of paired comparisons was undertaken on the assumption
that the Operations required to state which of two factors
was more influential were simpler than the operations
required to rate independently the influence of each
factor.. Attribution scores were derived by assigning one
point to a factor for each time it was preferred; the score
for each factor could thus range from 0 to 3.

All the dependent measures were read to each sub-
ject and each subject responded verbally. The responses

were scored by the experimenter.

Experimental task. The achievement task consisted

 

of two pictures, each containing eight hidden objects.

In an attempt to control Objective performance, three
Objects in each picture were made more salient so that all
children were able to find at least three Objects. Sub-
jects were asked to color in each hidden object they found.
Prior to beginning the task, subjects were told that they

would be timed. After finding three of the hidden objects,

35

the experimenter called "time." The subject was then told

that he did not find all the hidden Objects.

CHAPTER III

RESULTS

Achievement Judgment Stories

The first two hypotheses were concerned with how
the children would evaluate the hypothetical performances
of other children working a puzzle. These hypotheses were
tested by means of a 2 X 2 x 2 X 2 analysis Of variance
with subjects crossed with ability, effort, and outcome
and nested within grade level. The rewards and punish-
ments given to the achievement judgment stories were trans-
formed to a scale from one to ten with one being the most
punishment given (five red stars) and ten being the most
reward given (five gold stars). The analysis of variance
was performed on these transformed scores. Table 1 pre-
sents the means and standard deviations for the three vari—
ables by grade level.

The analysis Of variance revealed four significant
main effects and no significant interactions. Table 2
presents a partial summary of this ANOVA, as well as the

variance accounted for of each variable.

36

37

 

 

 

 

 

mm.ow mm.mn mm.mw mm.on m~.>m vm.mm
om.mauom om.mauom mo.mauom ¢~.wanom vv.mauom em.manom
Hm.an mm.~on : mm.omu z mm.mmu z mm.vnn z ma.o>n z vm.mhn z m opouo
m>.mauom me.bauom Ho.mauom H>.mauom mm.mauom mw.mauom
an.vw Hm.hmu z mw.abu z mm.mmu z am.mmu z mm.vmu z mo.mmu z a compo
zoo Ema: 30a amen 304 swam
oEoouso DH0mmm wuflaflnd

 

mouoom ucmempsb ucoam>mwnod mo mc0wumw>wa cumpcmum cam memo:

H magma

38

 

mmm vv.mhmmma HODOB

 

H>.m on mm.aom Houum

Hoo. sea. mm.H Ho.mma a Ho.mma < x O

Hoo. moH. v~.m Hm.mm~ a Hm.wm~ m x w

moo. Hoo. Ho.m mm.ooo H mo.oom o x o

moo. mmo. He.m Hm.mm¢ a Hm.mmv Ado muwawba

mmao. mooo. mm.mm o>.oomm a mo.ommm Amv DHOmmm
oma. mooo. Ha.mmm mo.mmhmm a mo.mm>mm Aoo GEOODDO

nmo. Hmo. om.v Hm.vmmn H Hm.vmmn ADV opouo

mocmauo> mo w .HMHcmHm m m: up mm mousom

 

<>Oz¢ mensmomz omummmom AmEooudoo N x Amuaaan<o N
x quooomo N x Iam>oq momuoo m x xuomflnsm mo xmmo m we» no sumEESm

N magma

39

AS Hypothesis 1 predicted, and as Table 2 indicates,
there was a main effect for outcome. Success was more
positively evaluated (X = 75.997) than failure (R = 60.234).

Hypothesis 2 stated, "among the older children the
amount of effort expended will be an influential evaluative
determinant." The grade X effort interaction did not reach
statistical significance as Table 2 indicates; however,
the trend, depicted by Figure l, is in the predicted
direction. While both first and third graders dispensed
rewards on the basis of expended effort, as indicated in
Table 2 by the main effect for effort, the mediating effect
of effort on rewards given was stronger for third graders
than for first graders.

Table 2 also indicates a main effect for ability,
which was not predicted. The variance accounted for by
this main effect, however, was only .0022.

Hypothesis 3 stated, "achievement judgments will
become more positive with increasing age." This hypothesis
received support from the main effect for age in the
analysis Of variance depicted in Table 2. An inspection
of the means for the two grade levels reveals that the
main effect is in the predicted direction. Third graders
gave more total rewards (i = 71.513) than did first

graders (x = 64.719).

40

18«P
16 4|— /
l4 «- o
12..
Number of 104-
Resultant Rewards
Allocated 84-

 

 

Grade Level
1

0-0 Outcome
A-A Effort
0-. Ability

Figure 1. Resultant Rewards Allocated for
Effort, Ability, and Outcome as
a Function of Grade Level.

41

Self Evaluations
The means and standard deviations of the self

evaluation responses are presented in Table 3.

Expectancy. Hypotheses 4 and 5 concerned the

 

children's expectancies for success assessed prior to doing
the task. These initial expectancies were subject to a 2
(Sex of Subject) X 2 (Grade Level) X 2 (Sex of Experimenter)
unweighted means analysis of variance. This ANOVA is sum-
marized in Table 4.

Hypothesis 4 stated, "younger children will have
higher expectancies than the older children." As Table 4
indicates, there is a significant main effect for grade
level. An inspection of the means for the two grade levels
reveals that the first—grade children had higher initial
expectancies (i = 6.675) than did the third-grade children
(R = 5.225). Additional support for Hypothesis 4 comes
from comparing the number of children at each grade who
indicated the highest possible choice (7) in response to
the initial expectancy question. Of 40 first grade
children, 33 chose 7 while only seven of the 40 third grade
children chose 7 on the expectancy scale.

Hypothesis 5 predicted that the third grade males
would have higher initial expectancies than would the third
grade females. An examination of Table 4 reveals a sex X
grade level interaction that approaches significance. This

trend is depicted in Figure 2. In the first grade, females

42

 

 

 

 

mm. now om. "om vm.uom mm. now oo.m "om Ho.auom vm.auom . Hm.anom
mm.mn z mn.au z om." z mm.au z om.m u z ma.au z oa.vu z mm.mn z m mpouu
mm. uom mm. Ham mm.uom om. now mn.~ Mom @O.Hnom om.auom mm. now
ma.mu z mm.Hu 2 mm." 2 mo.an z mm.HHu z mo.mn z mo.vu z mo.ou E H opmuo
Emma 5.. Mm he... some. some
mEOADDQfluuua monummmz ucopcmmoo

 

mmHDmmwz GOHUMDHMNKWIMHOW CO mGOflUMflNwmQ Chflmvcwuw GEM mammz

m magma

Summary of 2 (Grade) x 2 (Sex) x 2 (Sex of E)
ANOVA Performed on the Initial EXpectancies

43

Table 4

 

 

Source SS df MS F Signif.
Grade (G) 42.731 1 42.731 40.622 .001
Sex (S) .077 l .077 .073 .787
Sex of E (SE) 4.713 1 4.713 4.481 .038
G x S 3.025 1 3.025 2.876 .094
G x SE .109 1 .109 .104 .748
S x SE .014 l .014 .013 .910
G x S x SE .116 l .116 .110 .741
Error 75.737 72 1.052
Total 125.800 79 1.592

 

44

Initial 4a»
Expectancy

 

 

Grade Level
1 3

Female —---
Male ————

Figure 2. Initial Expectancies of Males and
Females as a Function of Grade Level.

45

have slightly higher initial expectancies than do males.
However, by third grade, this difference has been reversed.
Third grade males have slightly higher initial expectancies
than do third grade females.

Table 4 also indicates a main effect for sex of
experimenter. An inSpection of the means reveals that
expectancies were lower for male experimenters (5.875)
than for female experimenters (6.025).

Several of hypotheses 6 through 10 concerned the
attempt to manipulate attributions to ability. A manipu-
lation check reveals that the mean number of attributions
to ability did not differ for the ability instruction group
and the group not receiving any attribution orienting
instructions.

A 2 (Sex) X 2 (Grade Level) X 2 (Instruction
Condition) analysis of variance performed on the mean
number of attributions to ability shows that the attribution
instructions were not effective, independent of the age and
sex of the children. This ANOVA is summarized in Table 5.
Table 6 shows the mean number of attributions to ability
by grade and sex. Consequently, Hypotheses 6a, 7a, 8a,
and 9a cannot be tested since each depended on the efficacy
of the manipulation of attributions to ability. These
four hypotheses will not be examined in the following dis-

cussion.

Table 5

46

Summary Of 2 (Sex) x 2 (Grade Level) x 2 (Attribution
Instructions) ANOVA Performed on Attributions

to Ability

 

 

 

 

Source SS df MS F Signif
Sex (S) 1.012 1 1.012 1.378 .244
Grade (G) 2.812 1 2.812 3.828 .054
Instructions (I) .012 l .012 017 .897
S x G 1.012 1 1.012 1.378 .244
S x I .012 1 .012 .017 .897
G x I .612 1 .612 .834 .364
S x G x I 1.512 1 1.512 2.059 156
Error 52.900 72 .735
Total 59.888 79 .758
Table 6
Means and Standard Deviations Of Ability
Attributions by Grade and Sex
First Grade Third Grade
Male M 1.650 M 1.050 1.300
SD .875 SD .759
Female M 1.650 M 1.500 1.575
SD .933 SD .827
1.650 1.275

 

47

Hypothesis 6b stated that the expectancy drops of
the first grade children would not be related to the
attributions given following failure. Expectancy drops
were defined as the initial expectancy minus the final
expectancy. This hypothesis was examined by means of a
product-moment correlation for each of the four attributes
of ability, effort, task difficulty, and luck with the
expectancy drOpS. These correlations are presented in
Table 7. As predicted, neither of the two stable attributes
of ability and task difficulty were related to the expectancy
drops. Effort and luck were not related to the expectancy

drOps either. Thus Hypothesis 6b was confirmed.

Table 7

Correlations of Attributions with Expectancy DrOps

n...

 

 

Ability Effort Task Difficulty Luck

First Grade .145 -.0304 .017 -.120

Third Grade .189 -.304* .101 .022
*p < .05, two tailed test. N = 40

Hypothesis 6c predicted that the expectancy drOps
of the first graders would be larger than the expectancy
drOps of the third graders. To examine this hypothesis,
the expectancy drOps were subjected to a 2 (Sex) X 2 (Grade
Level) unweighted means analysis of variance. As can be

seen from Table 8, there was a significant difference in

48

Table 8

Summary of the 2 (Sex) x 2 (Grade Level) ANOVA
Performed on Expectancy Drops

m*-——-——— -_-___’ _. ———.—.—-—..~...~_-_ -o... -.—._‘ —4.— .—_._ -— 4——_.———< _.__.".—- 7 —.... - _. _. ._.._ .._.____
. _.-__-_.... __ -_. ‘-__.. .__.. _. -._.._. . __...-.....- - . .-...2_ .fi-.-

 

Source SS df MS F Signif.
Sex (S) .200 l .200 .074 .787
Grade (G) 16.200 1 16.200 5.956 .017
S x G .450 1 .450 .165 .685
Error 206.700 76 2.720
Total 223.550 79 2.830

 

the expectancy drops of the two age groups. An inspection
Of the means reveals that the difference was in the pre-
dicted direction. First graders dropped their expectancies
(R = 2.025) more than did third graders (R = 1.125).
Hypothesis 6c was confirmed.

Hypothesis 7b stated that the expectancy drOps
Of the third grade children would be related to the
attributions given following failure. This hypothesis was
examined by means of a product—movement correlation com-
puted for the four attributes of ability, effort, task
difficulty, and luck with the expectancy drops. As can be
seen from Table 7, only attributions to effort were
related to the drops in expectancies. Effort was, as pre-
dicted, negatively related to expectancy drops. None of

the other three attributes, however, were related to drops

49

in expectancy. Therefore, Hypothesis 7b received only
partiallsupport.

Hypothesis 7c predicted that the third grade
females would drop their expectancies more than would the
third grade males. A 2 (Grade Level) X 2 (Sex) unweighted
means analysis of variance performed on the expectancy
drops is summarized in Table 8. The predicted sex by grade
level interaction was not significant.

Hypothesis 7d stated, "Third grade females will
give more attributions to ability than will third grade
males." As the analysis of variance summarized earlier in
Table 5 shows, the predicted grade X sex interaction was
not significant.

Hypothesis 8 through 10 concerned the affective
responses of the children. To examine these hypotheses,
an affect index was computed by summing the response to
the two affective measures. This resulted in a scale from
2 to 14 with 2 being the most negative affect and 14 being
the most positive affect. Hypothesis 8b stated, "The
affective responses of the first grade children will not
be related to attributions following failure." This
hypothesis was examined by means of a correlation coefficient
computed for each of the four attributes of ability, effort,
task difficulty, and luck with the affect index. These
correlations are presented in Table 9. Contrary to

Hypothesis 8b, ability and luck were significantly

50

Table 9

Correlations for Attributions With Affect

 

Ability Effort Task Difficulty Luck
First Grade -.467** -.074 .180 .371*
Third Grade -.091 .043 .066 -.053

 

*p < .01, two tailed test.
**p < .001, two tailed test.

N = 40

correlated with the affective responses of the younger
children. The negative correlation for ability indicates
that as attributions to ability increased, negative affect
increased. The positive correlation between attribution to
luck and affect indicates that as luck attributions
increased, negative affect decreased. Both of these
relationships are consistent with Weiner's model.
Hypothesis 9b predicted that the affective response
of the Older children would be related to attributions made
following failure. Specifically, it was predicted that
greater negative affect would be expressed by those children
giving internal attributions, with attributions to ability
being particularly important. In addition, less negative
affect would be expressed by those children making external
attributions. As Table 9 indicates, none of the four

causal attributes were related to the affective responses

51

of the Older children. Hypothesis 9b, therefore, was not
confirmed.

Hypothesis 10 stated that the third grade children
would express more negative affect than would the first
grade children. This hypothesis was examined by means Of a
one-way analysis of variance performed on the affect index.
An examination of the mean affective reSponses of the two
grade levels reveals that the significant effect shown in
Table 10 was in the predicted direction. The affect index
for the first grade children was i = 11.375 and for the
third grade children the affect index was X = 9.800. The
means of the two age groups thus indicate that neither group
felt bad after failure since both groups achieved a mean
score above the neutral point of eight. Third grade
children therefore expressed less positive affect after

failure than did first grade children.

Table 10

Summary of One—Way (Grade Level) ANOVA Performed
on the Affect Ratings

 

....... - 4---... *—. -... -._—. ...- -—-__.__. ——-
*_ —. ...._.4 -———-——-—.—. .———.-._.- _.-..—___. .—

 

Source SS df MS F Signif.
Between Groups 49.613 1 49.613 6.674 .011
Within Groups 579.775 78 7.433

Total 629.388 79

 

52

Cognitive DevelOpment

The remaining four hypotheses concerned the impact
of concrete Operations on the mediating affect of causal
attributions on the achievement judgment and achievement
behavior of the first grade children. Recall that it was
assumed that the first grade children would be evidencing
an initial mastery of concrete Operations while the third
grade children would be established in concrete Operations.
This assumption is supported by examining the mean cognitive

task scores of the two grade levels. Out of a possible

eight points on the cognitive tasks, first graders achieved
a mean score of 4.175, while the third grade children
achieved a mean score of 6.900. The difference between
these two scores is significant (t = -5.51, df = 78,

p < .001).

Hypothesis 11 stated, "the mediating influence of
effort on achievement judgment will be related to the
children's level of cognitive develOpment." To examine
this hypothesis, a score representing the resultant of
the evaluation of effort minus the evaluation for lack of
effort was computed for the achievement judgment questions.
This score was then correlated with the total score on the
cognitive tasks. The correlation between this resultant
effort score and the cognitive score was .069.

In order to further examine this hypothesis, all

subjects were divided according to their stage of cognitive

53

development. Subjects were classified according to whether
they indicated preoperational thinking (scored 3 or less),
concrete (scored 7 or 8) or transitional thinking (scored
4, 5, or 6). The number of subjects in each category of
the above classification system is shown in Table 11. All
mean resultant effort scores were significantly different
from zero. A mean of zero would indicate that effort had
no influence on evaluations. Consequently, even the evalu-
ations of the preOperational children were influenced by

effort. Hypothesis 11 was clearly not confirmed.

Table 11

Mean Resultant Effort Score for Subjects at
Different Cognitive Levels

_.__. __

 

N Resultant Effort t df Signif.
Preoperational 13 14.615 3.665 12 .01
Transitional 35 5.457 2.848 34 .01
Concrete 32 5.688 4.490 31 .001

 

Hypothesis 12 predicted that the influence of
ability attributions on expectancy drops would increase
with increasing cognitive maturity. This hypothesis was
examined by means of a multiple correlation with ability
attributions and cognitive develOpment as the independent

variables and expectancy drops the dependent variable.

54

The zero-order correlations are presented in Table 12.

This hypothesis was not confirmed (R = .161).

Table 12

Zero-Order Correlations Used in Multiple Correlations
for Hypotheses 12 and 13

-. ,_.--- ---._ - .efi-.‘ »....-. ._... .».._... . ... -_.-_ - _ - ..- --_ .... o . a... - . . - - -...—
_... ‘4- «H- ..e .V M. - . .. .- -. - . _ I- . .— .. __ .. - _ -_... .. -.

 

CT AA
Cognitive Task (CT) X X
Ability Attributions (AA) -.050 X
Expectancy Drops (ED) .064 .145
Affect Index (AI) .051 -.467

 

N = 40

Hypothesis 13 predicted that cognitive develOpment
would be related to the influence of ability attributions
on affect. A multiple correlation was computed to examine
this hypothesis. Cognitive development and ability
attributions served as the independent variable with the
score on the affect index the dependent variable. The
zero-order correlations are presented in Table 12. The
multiple correlation turns out to be .468. Since the
multiple correlation is no higher than the zero-order
correlation for ability attributions and affect, Hypothesis
13 was not confirmed.

Finally, Hypothesis 14 stated, "high expectancies

of success will be negatively related to cognitive

55

development." The correlation between the score on the

cognitive tasks and the initial expectancy estimates of the

first grade children was .003. This final hypothesis was

not Confirmed.

CHAPTER IV

DISCUSSION

The findings which emerged from this study can be
considered under three general headings. First, three
hypotheses concerned the evaluative responses of children
to the achievement-related behaviors Of hypothetical others.
Second, several hypotheses focused on the children's
responses to failure. Finally, predictions were made about
the mediational processes--cognitive maturity and
attributions-—underlying these achievement-related evalu-
ations and responses.

With regard to the first point, the first three
hypotheses concerned the evaluative consequences of effort,
ability, and outcome in an achievement context. As pre-
dicted, outcome was the principal evaluative determinant.
Success was evaluated more positively than failure. In
addition, it was predicted that effort would be an important
determinant of evaluation only for the Older children.

While there was a tendency for the evaluations of the Older
children to be more influenced by effort, effort was an

influential evaluative determinant for both age groups.

56

57

This apparent discrepancy between the present study
and the Weiner and Peter (1973) investigation can best be
resolved by an examination of the age groups used in the
two studies. Weiner and Peter (1973), using age groups of
4-6 and 7-9, found that the evaluations of the 7-9 age
group, and not the 4—6 age group, were influenced by
effort. The present study, using age groups of 6-7 and 8-9,
suggests that by age seven effort has an important influence
on achievement-related evaluations. Consequently, were
Weiner and Peter (1973) to reanalyze their data using the
age groupings of 4-5, 6-7, and 8-9, the result would
probably support the results from the present investigation.

The unexpected finding of the influence of ability
on evaluations is not readily explained. Weiner and Peter
(1973) found that ability was an influential evaluative
determinant only for adults. Salili, Maehr, and Gillmore
(1976), however, found that ability had a significant
effect on the achievement evaluations of Iranian children.
These authors suggested that the divergence of their
results from the results of the Weiner and Peter study
(1973) was due to the critical role that culture is likely
to play in the development of achievement-related behaviors.
Specifically, Salili et a1. (1976) suggested that the
existence Of ability is more positively valued in Iran than
in the U.S. While it may be true that there are cultural

differences in the value placed on ability, it also seems

58

likely that the existence of ability will be differentially
valued within a culture. Consequently, the difference
between the results with regard to evaluation of ability

of the Weiner and Peter study (1973) and the present
investigation may be a result of different subject pOpu—
lations. The Weiner and Peter sample was drawn from a
large metropolitan area on the west coast, while the sample
for the present study was drawn from a small midwestern
community. Yet in order to fully examine subject popu-
lation differences data concerning the socioeconomic class
of individual subjects and achievement and IQ test scores
would be needed. Unfortunately, that data is not available
from either the Weiner and Peter study or the present
study. Finally, it should be noted that while the influ-
ence of ability on achievement evaluations in the present
study was statistically reliable, the variance accounted
for was quite small. In fact, as Table 1 indicated, out-
come clearly accounted for the most variance.

The findings with regard to the children's responses
to failure were generally consistent, with the notable
exception of certain predicted sex differences. The
finding that expectancies of success decrease as a function
of age is consistent with previous research (Parsons, 1974;
Parsons and Ruble, 1972). It is not clear from the present
research at what age expectancies no longer decrease as

age increases. Parsons (1974) found that children 9 1/2-11

59

years of age had lower expectancies than children 6 1/2-8
years of age. Unfortunately, most researchers have not
reported age differences when investigating expectancies
(McMahan, 1974; Crandall, 1969). Consequently, the age
range associated with lowered expectancies remains to be
determined.

Age differences were also found in reactions to
failure. First-graders dropped their expectancies more
following failure than did the third grade children.
Parsons (1974) also found, using age groups of 6 1/2-8 and
9 1/2-11, that the younger children's expectancies were
more influenced by failure than were the Older children's
expectancies. This may simply reflect the initial higher
expectancies of the younger children. In other words,
the larger expectancy drOps of the younger children may
partially reflect a regression effect. But if this were
true, the expectancies of the older children should be more
influenced by success. The expectancy increases of the
younger children should be attenuated by a regression
effect, whereas, the expectancy increases of the older
children would not reflect attenuation due to regression.
Such an analysis could not be examined in the present
study since success was not included as a variable. In
the Parsons study (1974), which did include success as a
variable, the expectancies of the younger children were

also more influenced by success. The expectancies of the

60

younger children increased more than did the expectancies
of the older children following success. This would sug-
gest that the expectancies of younger children are more
influenced by the outcome of an event, and not just failure.
It was expected that for the Older children,
females would have lower expectancies than males, which
would result in more ability attributions by females.
More ability attributions by females would in turn lead to
larger drOps of expectancy by females. This is essentially
the cycle postulated by Valle and Frieze (1976). None of
the sex differences suggested by the above analysis,
however, emerged in the present investigation. The critical
factor which initiates the above cycle is sex differences
in initial expectancies of success. It is not surprising,
therefore, that sex differences in ability attributions
and expectancy drops were not observed in the present study,
since no significant sex differences in initial expectancies
were found. McMahan (1973) provides support for the argu-
ment that sex differences in attribution patterns will
appear only when initial expectancies differ. In his study
(with sixth graders, tenth graders, and college students
as subjects), no differences in initial expectancies for
performance on an anagram test were found. As would be
expected, no differences between males and females on

attribution measures were found either.

61

The results of the present investigation and the
McMahan study (1973) bring into question the empirical
validity of the assertion that females consistently have
lower expectancies for success than males. Lenney (1977),
in an excellent review, suggests that there are at least
three situation variables that may affect women's per-
formance expectancies relative to men's. They include the
nature Of the task, clarity of performance feedback, and
the nature of certain social comparison cues. She con—
cluded that future research should be directed toward
empirically identifying the exact conditions in which
expectancies of both sexes increase or decrease.

Finally, as Parsons et a1. (1976) pointed out, age
is an important variable when considering sex differences
in expectancies. Sex differences have not been found prior
to the school age years.

The paper by Lenney (1977) suggests one possible
explanation of why sex differences in initial expectancies
were found in the study by Parsons (1974) but not in the
present investigation. Differences in the task must be
ruled out since a hidden figures task was used in both
studies. In addition, neither study emphasized social
comparison cues. The two studies did differ in that sub-
jects in the Parsons study (1974) did not receive any
feedback on the first trial. Expectancies were measured

before the second trial. In the present study, however,

62

expectancies were measured prior to the first trial. As
Lenney's review of the literature indicates, females Often
have lower expectancies than men when they are given minimal
or ambiguous feedback on their performances. Therefore,
the practice trial with no feedback may have contributed to
the lower expectancies of the females in the Parsons study
(1974). Finally, subjects in the Parsons study were all
tested by a female experimenter while half the subjects in
the present study were tested by male and half by a female
experimenter. While there was a main effect for sex of
experimenter indicating that lower expectancies resulted
from male experimenters, there was no interaction with

sex Of subject. Sex of the experimenter, therefore, cannot
account for the differences in the two studies.

With regard to the affective response of the
children, age again was an important factor. The affective
responses of the Older children were more influenced by
failure than were the affective responses of the younger
children. This finding is consistent with the results of
the Ruble et a1. (1976) study.

Since success or failure represents an evaluation
by the experimenter, it may be as Veroff (1969) suggests,
that younger children are not sensitive to the evaluations
Of others. "He [the preschool child] behaves in order to
bring about a condition of greater competence on his own.

Thus he evaluates on his own" (Veroff, 1969, p. 49). It is

63

only in the early school years that social evaluation
becomes important. Such an interpretation is consistent
with the present results. Furthermore, the age differences
in the affective responses are not readily explained within
an attributional framework. Since there were no age differ-
ences in the attributional patterns, there would be no
reason to expect age differences in affective responses.
The final group Of hypotheses concerned the medi-
ational processes underlying both the evaluation Of others
and responses to failure. First, it was predicted that
cognitive maturity would be related to the influence of
effort on achievement evaluations. The results of this
study, however, suggest that cognitive maturity, as measured
by Piagetian tasks, is not related to the children's
ability to use both effort and outcome cues in forming
evaluations. Thus while there may be stages and sequences
in the development of achievement judgments, as Weiner and
Peter (1974) suggest, such stages do not seem to be related
to cognitive development. The findings of Kun, Parsons,
and Ruble (1974) that even kindergarten children were able
to use both ability and effort cues to predict outcome is
consistent with this conclusion. The finding from the
present investigation that even preoperational children
were able to use both effort and outcome in forming evalu-

ations strongly suggests that the assumed central role of

64

cognitive maturity in the attribution process be recon-
sidered.

Further evidence against a central role for cog-
nitive maturity in the attribution process is suggested
by the failure to find a relationship between cognitive
development and the influence of ability attributions on
both expectancy and affect. Thus while the findings with
respect to the impact of concrete Operations on the medi-
ating influence of causal attributions were contrary to
predictions, they were at least consistent. Unfortunately,
the ineffectual attempt to manipulate attributions to
ability ruled out a more rigorous examination of the rela-
tionship between cognitive maturity and the attribution
process as it applies to achievement behavior. The failure
to find a global relationship does, however, suggest that
cognitive maturity beyond the preoperational stage is not
necessary for the emergence of the information-attribution-
affect and expectancy link found in adults. Thus, decen-
tration does not appear to be a unitary construct (Shantz,
1975).

Largely as a result of the assumptions concerning
the role of cognitive maturity in the attribution process,
it was predicted that attributions would mediate the
expectancy drops and affective responses of the older
children, and not the first grade children. But since

cognitive maturity was not related to the attribution

65

process, attributions should have influenced the expectancy
drops and affective responses Of both the first and third
grade children. Unfortunately, the results concerning the
influence of attributions were not consistent. For the
first grade children, internal attributions to ability
increased negative affect. This is consistent with previous
research by Nicholls (1975, 1976) who has found that ability
and not effort is related to affective responses. In
addition, external attributions to luck decreased negative
affect. This is consistent with the Weiner et a1. model
(1971). Yet attributions were not related to the expec-
tancy drOps of the younger children. In addition, attri-
butions were not related to the affective responses of the
Older children. Weak support was provided, however, for
the hypothesis that attributions would influence the
expectancy drops of the Older children. Only attributions
to effort were associated (negatively) with expectancy
drOps. This finding was also reported by Nichols (1975).
The inconsistency concerning the relationship
between causal attributions and expectancy and affect in
the present investigation may have been due to the untested
assumption that the four causal attributes of ability,
effort, luck, and task difficulty presented to the children
were the only important causes for their failure. In other
words, these four attributes may not be the ones children

readily use when confronted with a situation such as the

66

one in this experiment. Indeed, Frieze (1976) provides
evidence that adults may not use these four attributes, or
use others, when allowed to freely attribute causality.

She asked subjects to state why a particular event occurred.
Because the responses were entirely Open ended, subjects
were free to respond as they chose. The results revealed
that luck was seldom used and that mood, other people and
stable effort (lazy, energetic, etc.) were cited relatively
frequently, in addition to ability, effort and task diffi-
culty. Furthermore, 20% of the causal explanations given
by the subjects were unclassifiable. Taken together these
results suggest that a source of error may be introduced

by providing subjects with only the four attributes of
ability, effort, task difficulty, and luck to explain the
consequences of their behavior.

Furthermore, while adults may be able to reliably
categorize their causal explanations for events into the
four attributes listed above, it is less likely that
children, particularly young children, would be able to
generalize their own causal explanations to the four
attributes presented by the experimenter. Indeed, support
for this conclusion comes from the work of Brown and Berko
(1960) on semantic development. These researchers found
that the emergence of semantic generalization may not occur

until the end of the grade school years. In other words,

67

young children may not readily generalize to words of
related meaning.

The above analysis strongly suggests that children
may not be able to reliably utilize the four attributes of
ability, effort, task difficulty and luck used in the
present experiment, and most attribution studies, to
explain the results of their behavior. Consequently, an
important first step in studying the attributional processes
of children would be to determine the causal attributions
children actually use. A procedure similar to that used
by Frieze (1976) with adults would be one approach. More-
over, it may be unreasonable to assume that the same causal
categories are utilized by subjects in all situations.
Children particularly may use different causal attributes
in different situations. An examination of these questions
would seem imperative if consistent results are to be
expected from the study of children's attributional pro-

cesses .

Conclusion
The major purpose of this research was to examine
the role of cognitive maturity in the attribution process.
The results indicate, surprisingly, that children in the
preoperational stage of cognitive development are able to
use effort, ability, and outcome in forming evaluations of
the achievement behavior of hypothetical others. In

addition, the results indicate that cognitive maturity, at

68

least beyond the preoperational stage, does not mediate the
influence of causal attributions on either affect or expec—
tancies. Thus while achievement judgments and behaviors

do vary with age, these results suggest that cognitive
maturity does not determine or mediate such variation.
Further research is required to isolate the particular
variables which actually determine or mediate the variation
in achievement judgment and behavior with age.

Furthermore, as Wohlwill (1970) noted, correlational
procedures take on added significance when measuring change.
Most researchers examining the development Of the attri-
bution process have employed analysis of variance procedures
(Weiner & Peter, 1973; Ruble et al., 1976); however, this
may not be the most adequate procedure to follow. As
Wohlwill states, "The study of age changes entails a more
intensive concern with the measurement of change, and
methods for determining correlates of change and inter-
relationships among measures of change for different vari-
ables" (p. 57).

This study also indicates that while causal attri-
butions may have a significant effect on achievement judg—
ments and achievement behaviors, the variance accounted for
is quite small. Outcome accounted for ten times as much
variance as effort and ability combined in the achievement
judgments (.20 compared to .02). Moreover, the four causal

attributes of ability, effort, task difficulty, and luck

69

combined accounted for only .04 Of the first graders and .14
Of the third graders variance in expectancy drOps and .40
Of the first graders and .003 of the third graders variance
in affect. Thus further research is needed to identify
additional influences on achievement judgments and achieve-
ment behaviors. The influence of social evaluation (Veroff,
1969) may be one fruitful area of investigation.

Finally, this study offered evidence on the rela-
tionship between achievement judgments and achievement
behaviors. Third graders were more positive than first
graders in their evaluations of hypothetical others. Third
graders were less positive than first graders, however, in
their self evaluations. Unfortunately, a more rigorous
comparison of self and other evaluations was not possible
in this study. A study by McMahan (1975) does suggest,
however, that adults evaluate the achievement behavior of
others more positively than their own achievement behavior.
Additional research is needed to further examine the

relationship between self versus other evaluations.

APPENDICES

APPENDIX A

ACHIEVEMENT JUDGMENT STORIES

APPENDIX A

ACHIEVEMENT JUDGMENT STORIES

Paul is good at working puzzles. He is not trying to
do this puzzle. He does not get it put together.
What color star will you give Paul? How many?

Gold Stars

 

 

 

 

 

l 2 3 4 5

Red Stars

 

 

 

 

 

Carolyn is good at working puzzles. She is not trying
to do this puzzle. She gets it put together. What
color star will you give Carolyn? How many?

Gold Stars

 

 

 

 

l 2 3 4 5

Red Stars

 

 

 

 

 

7O

71

John is good at working puzzles. He is trying to do
this puzzle. He does not get it put together. What
color star will you give John? How many?

Gold Stars

 

 

 

 

 

1 2 a 4 5

Red Stars

 

 

 

 

 

Kathy is good at working puzzles. She is trying to do
this puzzle. She gets it put together. What color
star will you give Kathy? How many?

 

 

 

 

 

 

Gold Stars
—T_ _2’_ ‘7— 4 T
Red Stars
1 2 3 4 5

Tom is not good at working puzzles. He is not trying
to do this puzzle. He gets it put together. What
color star will you give Tom? How many?

Gold Stars

 

 

 

 

 

1 2 3 4 5

Red Stars

 

 

72

Pam is not good at working puzzles. She is not trying
to do this puzzle. She does not get it put together.
What color star will you give Pam? How many?

Gold Stars

 

 

 

 

 

1 2 3 4 5

Red Stars

 

 

 

 

 

Bob is not good at working puzzles. He is trying to
do this puzzle. He gets it put together. What color
star will you give Bob? How many?

Gold Stars

 

 

 

 

 

l 2 3 4 5

Red Stars

 

 

 

 

 

Paula is not good at working puzzles. She is trying
to do this puzzle. She does not get it put together.
What color star will you give Paula? How many?

Gold Stars

 

 

 

 

 

l 2 3 4 5

Red Stars

 

 

 

 

 

APPENDIX B

INSTRUCTIONS FOR HIDDEN FIGURES TASK

APPENDIX B

INSTRUCTIONS FOR HIDDEN FIGURES TASK

Hidden Figures Instructions-—Ability

I've got some pictures here that have animals
hiding in them. I want to see if you can find all the
animals that are hiding in this picture. When you see one
Of the animals, I want you to color its face with this
crayon.

Now, we've found that some children are better at
finding the animals than other children. Some children
are just good at finding the hidden animals. It doesn't
make much difference how hard you look. Why do some
children find all the hidden animals? (Be sure child
understands that some children are just better at the task.)
Some children just seem to be better at finding the hidden
animals than other children. It doesn't make much differ-
ence whether a child looks very hard or takes it easy.

SO if you find all the animals that are hiding in the
picture, it is because you are just good at finding hidden
animals. SO, at the end of the time limit we'll see if
you found all of the hidden animals.

Hidden Figures Instructions--Control

I've got some pictures here that have animals
hiding in them. I want to see if you can find all the
animals that are hiding in this picture. When you see one
of the animals, I want you to color its face with this
crayon. SO, at the end of the time limit we'll see if
you found all of the hidden animals.

73

APPENDIX C

DEPENDENT MEASURES FOR HIDDEN FIGURES TASK

APPENDIX C

DEPENDENT MEASURES FOR HIDDEN FIGURES TASK

Child's Questionnaire

1a. This picture shows how good seven children your age

were at finding the hidden animals. This child was
the best at finding the animals (point to figure under
the tallest bar) and this child was the worst at
finding the animals (point to the figure under the
shortest bar). SO the taller the bar the better the
child did. Point 3 & 4 and say: Look at these two
children and tell me which one is better at finding
the hidden animals (explain again if child doesn't
understand). Now then I want you to show me which
child you think you will turn out to be.

1 2 3 4 5 6 7
best worst

This shows how much seven children liked playing this
game. This child (point to figure under tallest bar)
liked looking for the hidden animals the most. And
this child (point to figure under shortest bar) didn't
like looking for the animals at all. SO, the taller
the bar the more the child liked looking for the
hidden animals. Look at these two children (point to
5 & 6) and tell me which one liked looking for the
animals more. (Make sure the child understands the
progressive ordering.) Now, I want you to show me
which you turned out to be.

1 2 3 4 5 6 7
didn't like did like

74

3.

75

I want to know why you think you didn't find all the
animals that were hiding in the picture. Was it
because:

you didn't try hard OR you weren't lucky

the animals were not easy to find
OR
you are not good at finding animals

you were not lucky
OR
the animals were not easy to find

you are not good at finding animals hiding in
pictures

OR
you did not try hard

you were not lucky

OR
you are not good at finding animals hiding in
pictures

you did not try hard
OR
the animals were not easy to find

Let's look at this picture again. This time it shows
how happy seven children your age were about the number
of animals they found in the picture. This child was
very happy about the number of animals he found (tallest
bar) and this child was very sad about the number of
animals he found (shortest bar). SO, the taller the

bar the happier the child is. Look at these two
children (2 & 4) and tell me which one is happier about
the number of animals he found. Now, I want you to

show me which child you turned out to be.

1 2 3 4 5 6 7
sad happy

76

This picture shows how good seven children your age
were at finding the hidden animals. This child was
best at finding the animals (point to figure under
tallest bar) and this child was the worst at finding
the animals (point to figure under the shortest bar).
SO the taller the bar the better the child. Point to
3 & 4 and say: Look at these two children and tell me
which one is better at finding the hidden animals
(explain again if the child doesn't understand). Now
then I want you to Show me which child you think you
turned out to be.

1 2 3 4 5 6 7
best worst

LIST OF REFERENCES

LIST OF REFERENCES

Battle, E. Motivational determinants of academic task
persistence. Journal of Personality and Social
Psychology, 1965, 2, 259-218.

 

 

Brim, O. G. Jr., Goslin, D. A., Glass, D. C., & Goldberg, I.
American Beliefs and Attitudes about Intelligence.
New York: Russell Sage, 1969.

Brown, R., & Berko, J. Word association and the acqui-
sition of grammar. Child Development, 1960, 31,
1-14 0

 

Crandall, V. C. Sex differences in expectancy of intel-
lectual and academic reinforcement. In C. P.
Smith (ed.). Achievement-related Motives ip
Children. New York: Russell Sage, 1969.

 

Dweck, C. S. The role of expectations and attributions in
the alleviation of learned helplessness. Journal
of Personality and Social Psychology: 1975, 32,
1577-1088. __

Dweck, C. S., & Bush, E. 8. Sex differences in learned
helplessness: 1. Differential debilitation with
peer and adult evaluators. Developmental
Psychology, 1976, 12, 147-1561

 

 

Dweck, C. S., & Gilliard, D. Expectancy statements as
determinants of reactions to failure: Sex differ-
ences in persistence and expectancy change.

Journal pf Personality and Social Psychology,
Igjsr 23: 1 " '

Dweck, C. S., & Reppucci, N. D. Learned helplessness and
reinforcement responsibility in children. Journal
of Personality and Social Psychology, 1973, 25,
I694116.

 

Elkind, D. Children's discovery of the conservation of
mass, weight, and volume: Piaget replication study
II. Journal pf Genetic Psychology, 1961a, 98,
219-227.

 

77

78

Elkind, D. The development of the additive composition
of classes in the child: Piaget replication study
III. Journal gt Genetic Psychology, 1961b, 99,
51-57. .

 

Feather, N. T. Effects of prior success and failure or
expectations of success and subsequent performance.

Journal pt Personality and Social Psychology, 1966,
9, 287—298.

 

Feather, N. T. Attribution of responsibility and valence
of success and failure in relation to initial con-
fidence and task performance. Journal of
Personality and Social Psychologxz 19697—19,
129-144.

 

 

Feather, N. T., & Simon, J. G. Attribution of respon-
sibility and valence of outcome in relation to
outcome in relation to initial confidence and suc-
cess and failure of self and other. Journal of
Personality and Social Psychology, 1971a, 99,——
173-188.

 

Feather, N. T., & Simon, J. G. Causal attributions for
success and failure in relation to expectations of
success based upon selective or manipulative
control. Journal gt Personality, 1971b, 99,
527-541.

 

Feather, N. T., & Simon, J. G. Luck and the unexpected
outcome: A field replication of laboratory findings.
Australian Journal gt Psychology: 1972, 91, 113-
117.

 

 

Feffer, M., & Gourevitch, V. Cognitive aspects of role-
taking in children. Journal pt Personality, 1960,
99, 383-396.

 

Flavell, J. H. The Developmental Psychology gt Jean Piaget.
D. Van Nostrand Company, 1963.

 

 

Frieze, I. H. Causal attributions and information seeking
to explain success and failure. Journal gt Research
13 Personality, 1976, 99, 293-305.

 

Frieze, 1., & Weiner, B. Cue utilization and attri-
butional judgments for success and failure.
Journal gt Personality, 1971, 99, 591-606.

 

79

Hill, K. T., & Dusek, J. B. Children's achievement expec-
tations as a function of social reinforcement, sex
of subject, and text anxiety. Child Development,
1969, 49, 547-557.

 

Inhelder, B., & Piaget, J. The Growth gt ngical Thinking
from Childhood t9 Adolescence. New York: Basic
Books, 1958.

 

 

 

 

Kun, A., Parsons, J. B., & Ruble, D. N. DevelOpment of
integration processes using ability and effort
information to predict outcome. Developmental
Psychology, 1974, 99, 721-732.

 

 

Lee, L. C. The concomitant development of cognitive and
moral modes of thought: A test of select deductions
from Piaget's theory. Genetic Psychology Mono-
graphs, 1971, 99, 93-146.

 

Lenney, E. Women's self-confidence in achievement settings.
Psychological Bulletin, 1977, 94, 1-13.

 

McMahan, I. D. Relationships between causal attributions
and expectancy of success. Journal pf Personality
and Social Psyghology, 1973, 29, 108-114.

 

 

McMahan, I. D. Sex and person in achievement and moral
evaluation. Developmental Psychology, 1975, 9;,
659-660.

 

Montanelli, D. S., & Hill, K. T. Children's achievement
expectations and performance as a function of two
consecutive reinforcement experiences, sex of
subject, and sex of experimenter. Journal gt
Personality and Social Psychology, 1969, i9, 115-
128.

 

Nichols, J. Causal attributions and other achievement
related cognitions: Effects Of task, outcome,
attainment value and sex. Journal of Personality
and Social Psychology, 1975, 9;, 3793389.

 

 

Nicholls, J. G. Effort is virtuous, but it's better to
have ability: Evaluative reSponses to perceptions
of effort and ability. Journal pt Research 33
Personality, 1976, $9, 306-315.

 

Parson, J. E., & Ruble, D. N. Attributional processes
related to the development of achievement-related
affect and expectancy. Paper presented at the
meeting of the American Psychological Association,
Honolulu, August 1972.

8O

Parsons, J. E., Ruble, D. N., Hodges, K. L., & Small,

Phares,

Piaget,

Piaget,

Piaget,

Piaget,

Piaget,

Riemer,

Rotter,

Rotter,

Ruble,

Salili,

Selman,

A. W. Cognitive-develOpmental factors in emerging
sex differences in achievement-related expectancies.
Journal 9: Social Issues, 1976, 99, 47-61.

 

E. J. Expectancy changes in skill and chance
situations. Journal ot Abnormal and Social
Psychology, 1957, 91, 339-342.

 

 

 

 

 

J. The Language and Thought ot the Child. New
York: Harcourt, Brace, l .

J. Judgement and Reasoning to the Child. New
York: Harcourt, Brace, 1928.

 

 

J. The Child's Conception o: the World. New York:
Harcourt, Brace, 1929.

 

 

J. The Child's Conception 9: Number. London:
Routledge, 1952.

 

J. The Moral Judgement o: the Child. Translated
by M. Gabain. New York: Free Press, 1965.

 

 

B. S. The influence of causal beliefs on affect
and expectancy. Journal of Personalit and
Social Psychology, 1975, :1, - .

 

J. B. Generalized expectancies for internal
versus external control of reinforcement.
Psychological Monographs, 1966, 99, 1 (Whole #609).

 

J. B., Liverant, S., & Crowne, D. P. The growth
and extinction of expectancies in chance controlled
and skilled tasks. Journal gt Psychology, 1961,
99, 161-177.

 

D. N., Parsons, J. E., & Ross, J. Self-evaluative

responses of children in an achievement setting.
Child Development, 1976, 91, 990-997.

 

F., Maehr, M. L., & Gillmore, G. L. Achievement
and morality: A cross-cultural analysis of causal

attribution and evaluation. Journal of Personality
and Social PsycholOQY, 1976, 99, 327-337.

 

 

R. The relation Of role taking to the development
of moral judgement in children. Child Development,
1971, 59, 79-91.

 

81

Shantz, C. U. The develOpment of social cognition. In
E. M. Heatherington (ed.). Review gt Child
DevelOpment Research (Vol. 5). UniverSIty of
Chicago Press, 1975.

 

Simon, J. G., & Feather, N. T. Causal attributions for
success and failure at University examinations.
Journal gt Educational Psychology: 1973, 91, 46-56.

 

Tomlinson-Keasey, C., & Keasey, C. B. The mediating role
of cognitive develOpment in moral judgement.
Child Development, 1974, 19, 291-298.

 

Valle, V. A., & Frieze, I. H. Stability of causal
attributions as a mediator in changing expec-

tations for success. Journal of Personality and
Social Psychology: 1976, 99, $79—$87.

 

 

Veroff, J. Social comparison and the develOpment of
achievement motivation. In C. P. Smith (ed.),
Achievoment-related Motives to Children. New
York: Russell Sage, I969.

 

Weiner, B. Theories o9 Motivation: From Mechanism to
Cognition. Markham, 1972.

 

 

 

 

Weiner, B. Achievement motivation as conceptualized by an
attribution theorist. In B. Weiner (ed.).
Achievement Motivation and Attribution Theoty.
General Learning Press, 1974.

 

Weiner, B., Frieze, I., Kukla, A., Reed, L., Rest, S., &
Rosenbaum, R. M. Perccivingthe Causes of Success
and Failure. General Learning Press,71971.

 

 

Weiner, B., & Kukla, A. An attributional analysis of
achievement motivation. Journal 9: Personality
and Social Psychology, 1970, 99, 1-20.

 

 

Weiner, B., Nierenberg, R., & Goldstein, M. Social learn-
ing (locus of control) versus attributional (causal
stability) interpretations of expectancy of success.
Journal of Experimental Social Psychology, 1976,
1'2—52—_, 49':

 

Weiner, B., a Peter, N. A cognitive-developmental analysis
of achievement and moral judgements. DevelOpmental
Psychology, 1973, 9, 290-309.

 

 

Wohlwill, J. F. The age variable in psychological research.
Psychological Review, 1970, 11, 49-64.

 

DIFFERENT METHODOLOGICAL APPROACHES TO
PHILOSOPHY OF SCIENCE