The Pictorial Scale of Perceived Competence and Social Acceptance for Young Children

Author(s): Susan Harter and Robin Pike

Source: Child Development , Dec., 1984, Vol. 55, No. 6 (Dec., 1984), pp. 1969-1982
Published by: Wiley on behalf of the Society for Research in Child Development

Stable URL: https://www.jstor.org/stable/1129772

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Society for Research in Child Development and Wiley are collaborating with JSTOR to digitize,
preserve and extend access to Child Development

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
The Pictorial Scale of Perceived Competence
and Social Acceptance for Young Children

Susan Harter and Robin Pike

University of Denver

HARTER, SUSAN, and PIKE, ROBIN. The Pictorial Scale of Perceived Competence and So
tance for Young Children. CHILD DEVELOPMENT, 1984, 55, 1969-1982. A new pictorial
ceived competence and social acceptance for young children, a downward extension
ceived Competence Scale for Children, is described. There-are 2 versions of this instru
preschoolers and kindergartners and a second for first and second graders, each ta
mains: cognitive competence, physical competence, peer acceptance, and maternal
Factor analyses reveal a 2-factor solution. The first factor, general competence, is def
cognitive and physical competence subscales. The second factor, social acceptance, co
peer and maternal acceptance subscales. The psychometric properties were found to be
Weak correlations between children's and teachers' judgments are discussed in terms o
child's tendency to confuse the wish to be competent or accepted with reality. It is urg
instrument not be viewed as a general self-concept scale but be treated as a measure of
constructs, perceived competence and social acceptance.

Introduction perceptions across these domains. This ap-

proach is based on the assumption that chil-
Constructs such as self-concept and self-
dren do not view themselves as equally ade-
esteem have had a long history within the
quate in all domains; rather, we have assumed
field of psychology, although in recent years
that they are capable of making meaningful
there has been a revival of interest in topics
distinctions between different domains.
involving the self and self-description (see
Harter, 1983; Harter, in press). Despite thisSupport for this assumption has been ob-
tained
interest, relatively little attention has been de- in measurement efforts with children
voted to the sensitive measurement of such and adolescents between the ages of 8 and 18.
In the construction of the Self-Perception
constructs, particularly in the young child (see
also Wylie, 1979). The present article de- Profile for Children (Harter, Note 1), we dem-
scribes the construction of a new scale de- onstrated that children differentiate the fol-
signed to assess perceived competence and
lowing five domains as revealed through
factor-analytic procedures: scholastic com-
social acceptance in young children, ages 4-7.
petence, athletic competence, ,'eer accep-
Our conceptual approach to the assess-
tance, physical appearance, and conduct or
ment of self-judgments has been domain-
behavior. In addition to judgments in these
specific, unlike the frameworks adopted by
specific domains, children aged 8 and older
other test constructors (e.g., Coopersmith,
can also make reliable judgments about their
1967; Piers & Harris, 1969), who have sought
general worth as a person. Thus, the current
to assess self-concept or self-esteem primar-
structure of the Self-Perception Profile for
ily through the calculation of a single score,
older children contains five separate domain-
summing items across diverse domains. In
specific subscales as well as a sixth subscale
contrast, we have sought to assess children's
tapping self-worth.
self-judgments separately within specific
domains in order to provide a profile of self- In more recent efforts to devise a self-

The research necessary for the development of this scale was supported by grant HD-09613
from the National Institute of Child Health and Human Development of the Department of Health,
Education, and Welfare. The authors would like to acknowledge the extensive cooperation of both
the school personnel and pupils from the following school systems, without whose assistance this
scale could not have been constructed: the Cherry Creek, Denver, and Jefferson County public
school systems, the Jewish Community Center of Denver, the Evergreen Children's Center, St.
Timothy's Preschool, Wellshire Preschool, and Parker Preschool. The authors also thank the research
personnel in our group, who worked extensively on the development of the scale, including Carole
Efron, Christine Chao, and Beth Ann Bierer. Reprints and information on obtaining materials can be
obtained by writing to Susan Harter, Department of Psychology, 2040 S. York Street, Denver,
Colorado 80208.

[Child Development, 1984, 55, 1969-1982. ? 1984 by the Society for Research in Child Development, Inc.
All rights reserved. 0009-3920/84/5506-0020$01.00]

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
1970 Child Development

report instrument for young children,butwe more scholastically oriented skills such as
adopted a similar approach in that we soughtbeing able to spell, read, or add are better
measures of cognitive competence in the first
to identify meaningful domains in the child's
life and to construct separate subscalesand forsecond grades.
each. We also opted for utilization of a similar
The younger children's instrument also
type of question format that (a) provides a from the older version in that it con-
differs
greater range of responses for each item (four
tains no self-worth subscale. Both theory (see
choices rather than the more typical two-
Harter, 1983) and empirical findings have led
choice true/false format) and (b) reduces chil-
to the conclusion that children are not capable
dren's tendency to give the socially desirable
of making judgments about their worth as per-
response (see Harter [1982] for a more com-
sons until approximately the age of 8. The
plete description).
very concept of "personness" is not yet firmly
However, in devising a developmentally established among younger children, nor is
appropriate downward extension of the scale the notion that the self, so defined, can be
evaluated as a global entity.
for 4-7-year-olds, several aspects of our proce-
dure were different from those used with the
There is another developmental contrast
older children. First, a pictorial format was de-
that involves the degree to which we can ex-
vised rather than a written questionnaire. Ex-
pect children's self-judgments to be accurate.
perience shows that, on questionnaires, young
Developmental frameworks such as those of-
children's inability to read as well as to under-
fered by Piagetians or the proponents of
stand the items, coupled with related atten-
psychoanalytic theory would alert us that the
tional problems, attenuates both the reliability
judgments of the young child may not be real-
and validity of such instruments. In contrast,
istic. That is, young children confuse the wish
the pictorial format engages the young child's
to be competent with reality; they blur the
interest, is understandable, sustains the distinction between their ideal self-image and
child's attention, and leads to more meaning-
the real self (Stipek, 1981). Related findings
ful responses.
by Ruble and her colleagues (see Ruble, 1983)
The pictorial format also allows us to de-indicate that it is not until approximately 9
years of age that children make use of social
pict skills and specific activities concretely.
comparison for the purposes of judging their
Whereas, at older ages, trait labels and general
descriptions of skill or adequacy can be em- own competence. Thus, certain cognitive lim-
ployed-such as terms like smart, popular, itations appear to interfere with the young
child's ability to make realistic judgments
athletic, and good-looking-the young child
about the self.
has not yet acquired these forms of self-
description (see Harter [1983] for a theoretical
discussion of these developmental shifts). Given that young children may not be
Rather, the young child's self-judgmentsvery
in- accurate judges of their competence or
social acceptance, comparisons of their scores
volve the behavioral description of their
specific abilities, such as completing puzzles,
with objective indexes should not be exam-
ined as an index of the validity of the instru-
running fast, and playing with friends. There-
ment. This lack of convergence is an inter-
fore, the graphic presentation of these actions
and activities facilitates the young child'sesting finding, in and of itself, one that bears
on the self-descriptive capabilities during this
understanding of the task since these forms of
developmental period. Other forms of valid-
self-description are developmentally appro-
priate.
ity-such as discriminant, convergent, and
predictive validity-would appear to be more
Another difference involves the number appropriate, as will be demonstrated.
of versions of the scale required. At older
ages, one version can be utilized across a wide The specific content of the scale to be
range of ages. For the younger ages, however, described involves two general constructs,
it was necessary to devise one version for pre- perceived competence and perceived social
schoolers and kindergartners (4- and 5-year- acceptance. The measure contains two
olds) and a separate version for first and sec- subscales within each of these domains. Per-
ond graders (6- and 7-year-olds). This was ceived competence is divided into two sub-
necessitated by the fact that the specific skills scales, cognitive competence and physical
that define or connote competence and social competence. Social acceptance is divided into
acceptance change rather dramatically within two subscales, peer acceptance and maternal
this 4-year age range. For example, puzzles acceptance. While these particular subscales
may be indicative of cognitive competence appear to define salient domains in the life of
during the preschool and kindergarten years, the young child, obviously there are others,

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
 Harter and Pike 1971

three of which (paternal acceptance,

dren teacher
were white, with the remaining 4% His-
acceptance, and conduct) will appear inblack,
panic, sub- and Oriental.
sequent versions of this scale.
Scale Description
Given the structure of this scale, we
Scale structure.-The scale contains four
strongly urge that the scale not be viewed
separate as
subscales-cognitive competence,
physical
an index of self-concept or self-esteem percompetence,
se. peer acceptance, and
That is, certain judgments such as maternal
the percep- acceptance. Each subscale contains
tion of one's cognitive or physical sixcompetence
items. There are two versions of the scale,
or one's behavior may well reflect one judgments
for preschool-kindergarten and one for
about the self's capabilities. However,
first andper- second grades. These two versions
ceptions concerning the degree toare which one
not completely unique. Rather, for certain
has friends or obtains support from parentsthere
subscales, or are overlapping or common
teachers do not necessarily imply items judgments
across the two versions. Table 1 pro-
about the adequacy of the self. For example,
vides a master list of all items for each version,
one may conclude that something grouped about the according to subscale. Asterisks next
self is responsible for one's lack of to the item On
friends. number indicated which items are
the other hand, the cause may reside common in to both forms.
cer-
tain characteristics of one's peers; that is, they
As can be seen in Table 1, none of the
are not nice or are not friendly. Similarly, lack
items that define the cognitive subscales at
of parental support might be because one per-
the two developmental levels overlap. The
ceives the self as unlovable, yet, on the other
preschool-kindergarten form contains a num-
hand, one may perceive one's parents as un-
ber of rudimentary readiness skills (knowing
loving. Thus, the very basis on which chil-
colors, the alphabet, being able to count) in
dren make such judgments is an interesting
addition to performance on puzzles and ob-
empirical issue itself. However, until further
taining stars on papers. The first-second grade
research has clarified this issue, we would
version do those scholastic skills ini-
includes
well not to assume that all of these seeming
tially encountered in the early primary grades
self-judgments are based on characteristics
(reading, writing, and arithmetic).
that reside in the self. For this reason we have
For theof
urged that the scale be treated as a measure domain of physical skills, four
items
what the title indicates, perceived occur on both versions (swinging,
compe-
climbing,
tence and perceived social acceptance, skipping, and running). Two of the
rather
preschool-kindergarten
than treating it as a singular measure of "self- skills (tying shoes and
concept" or "self-esteem." hopping), however, are replaced by more ad-
vanced physical skills for the first-second
Method grade version (bouncing a ball and jumping
rope).
This particular instrument has undergone
numerous revisions in terms of scale struc- Within the domain of peer acceptance,
ture, item content, and question format, based four of the items involving friends are com-
on extensive piloting with large numbers ofmon across the two versions. Two of the items
subjects. In this article, we will restrict ouron the preschool-kindergarten version (stay-
description to the final version of the four-ing overnight and eating at friends' houses)
subscale instrument. are replaced at the first-second grade level by
others sharing toys and others sitting next to
Subjects
you. These particular social overtures in the
Subjects were 90 preschoolers (mean age
early primary grades would appear to be im-
= 4.45), 56 kindergartners (mean age = 5.54),
portant indexes of popularity.
65 first graders (mean age = 6.32), and 44 sec-
ond graders (mean age = 7.41), approximately For the domain of maternal acceptance,
equally divided by gender within each group. there are four activities or maternal behaviors
These samples provided the primary data for in common across the two age-graded ver-
the factor analyses, means, standard devia- sions (Mom takes you places you like, cooks
tions, internal consistency reliability data, and "your favorite foods, reads to you, and talks to
subscales intercorrelations reported. For an you). Two preschool-kindergarten items drop
additional sample of 77 preschoolers, 28 kin- out (Mom smiles and Mom talks to you) for the
dergartners, and 38 first and second graders, first-second grade version and are replaced by
scores for both self-report and teacher ratings Mom lets you eat at friends' and stay over-
were available. All subjects were drawn from night. These maternal acceptance items were
schools in middle-class neighborhoods. In generated from a list of the most commonly
terms of ethnic composition, 96% of the chil- mentioned behaviors by young children in re-

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
1972 Child Development
TABLE 1

ITEMS GROUPED ACCORDING TO SUBSCALE FOR EACH FORM

Subscale and Item No. Preschool-Kindergarten First-Second Grades

Cognitive competence:
1 ................... Good at puzzles Good at numbers
5 .................... Gets stars on paper Knows a lot in school
9 .................... Knows names of colors Can read alone
13 .................... Good at counting Can write words
17 ..................... Knows alphabet Good at spelling
21 ................ ..... Knows first letter of name Good at adding
Physical competence:
3 ..................... Good at swinging Good at swinging
7* ..................... Good at climbing Good at climbing
11 ................... Can tie shoes Good at bouncing ball
15* .................. Good at skipping Good at skipping
19* ..................... Good at running Good at running
23 ................... Good at hopping Good at jump-roping
Peer acceptance:
2* .................. Has lots of friends Has lots of friends
6 ................... Stays overnight at friends' Others share their toys
10* .................. .. Has friends to play with Has friends to play with
14* ................. .. Has friends on playground Has friends on playground
18* ..................... Gets asked to play with others Gets asked to play with othe
22 ..................... Eats dinner at friends' house Others sit next to you
Maternal acceptance:
4 ................... Mom smiles Mom lets you eat at friends'
8* .................. Mom takes you places you like Mom takes you places you like
12*" ........................ Mom cooks favorite foods Mom cooks favorite foods
16* ....................... Mom reads to you Mom reads to you
20 .................... Mom plays with you Mom plays with you
24* .................. Mom talks to you Mom talks to you

NOTE.-Item number refers to position of the item in the order administered to t

items common to both forms.

sponse to the question, "Tell me thethe

ple item, things
female subject would be told that
the girlthat
your mother does that let you know on theshe
child's left is good at puzzles
likes or loves you." but the child on the right is not very good at
puzzles. The child's first task is to indicate
Picture plates.-The pictures accom-
which of the two girls she is most like. After
panying each version are bound separately, as
making that decision, the child is then asked
are sets for boys and girls. Thus, there are four
to think only about the picture on that side
books of plates, both a boys' and girls' set for
and indicate whether she is a lot like that girl
the preschool-kindergarten and first-second
(the big circle) or just a little bit like that girl
grade versions. The activities depicted in
(the smaller circle). For each item there are
each item are identical for girls and boys.
more specific descriptive questions that ac-
Only the gender of the target child is differ-
company each circle, such as "Are you just
ent, so that a subject can respond to pictures
pretty good at puzzles [small circle] or really
depicting a same-gender child.
good [large circle]?" The book of plates is
Items occur in the order of cognitive com- constructed so that, as the picture for a given
petence, social acceptance, physical compe- item is presented to the child, the item de-
tence, and maternal acceptance, and continue scription to be read by the examiner sitting
to repeat themselves in that order. Withinopposite the child is printed on the back of the
each subscale, items are counterbalanced sopreceding picture (see manual [Harter & Pike,
that three of the pictures depict the most com- Note 2] for more specific instructions).
petent or accepted child on the left and three
Scoring.-Each item is scored on a four-
of them depict the more competent or ac-
point scale, where a score of 4 would be the
cepted child on the right.
most competent or accepted and a score of 1
Sample item.-The scale is individually would designate the least competent or ac-
administered. A sample item is presented in cepted. Thus, for the sample item, the child
Figure 1. The child is first read a brief state- who indicates that she is a lot like the girl on
ment about each child depicted. For the sam- the left who is good at puzzles would receive

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
 Harter and Pike 1973

FIG. 1.-Sample item

a score of 4. If she chose the smaller circle on Subscale reliabilities, in the form of internal
the left, she would get a 3. If she indicates that consistency coefficients, will then be pre-
she is a little like the girl on the right who is sented, followed by subscale means and stan-
not very good at puzzles, she would receive a dard deviations. Intercorrelations among sub-
2. And if she is a lot like that girl, she would scales, as well as correlations between child
get a score of 1. (These scores are designated and teacher ratings, will then be described.
on a scoring key under the verbal descriptions
provided for the examiner for each item in the Factor Pattern
picture plates.) Item scores are averaged Tables 2 and 3 present the factor pattern
across the six items for a given subscale, and based on an oblique (promax) rotation, a solu-
these four means provide the child's profile of tion that allows the factors to intercorrelate.
perceived competence and social acceptance. This solution was considered the most appro-
priate given our expectation, based on previ-
Teacher rating scale.-A teacher rating
ous findings, that there would be moderate
scale parallels the child's instrument.
Teachers rate the child in three of the four and meaningful correlations among self-
judgments in these domains. Cattell's "scree"
areas tapped on the child's version: cognitive
text, based on the magnitude of the eigen-
competence, physical competence, and peer
values, as well as interpretability, indicated
acceptance. (We did not feel that it was appro-that a two-factor solution best described the
priate to have teachers rate the maternal ac-
data from both the combined preschool-
ceptance of the child.) On this scale, teachers
kindergarten samples as well as the combined
are given a brief verbal description of each
first-second grade samples.'
item (e.g., good at puzzles, has lots of friends,
good at swinging) and then rate how true that As can be seen in Tables 2 and 3, for both
statement is on a four-point scale (really true,
groups, items generally have moderate to high
pretty true, only sort of true, and not very loadings on their designated factor, and with
true). Thus, these scores can be comparedtwo exceptions for the preschool-kindergarten
with the child's scores, depending on the pur-
sample, items do not cross-load on the other
poses of the study. factor. Loadings are somewhat higher for the
Results
first-second grade samples. (Loadings less
than .19 are not presented, for the sake of clar-
The primary results bear upon the ity.) Factor 1 is defined by the two compe-
psychometric properties of the scale. To de- tence subscales, cognitive and physical; thus
termine the factorial validity of the scale, theit is considered to reflect perceptions of gen-
factor pattern will first be presented, along eral competence. Factor 2 is defined by the
with item means and standard deviations. peer acceptance and maternal acceptance sub-

1 Initially we performed the more traditional orthogonal rotation, which also revealed a tw
factor solution. However, the oblique rotation not only seemed more appropriate conceptually b
provided a somewhat better fit.

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
1974 Child Development
TABLE 2

FACTOR PATTERN AND ITEM MEANS AND STANDARD DEVIATIONS FOR THE PRESCHOOL AND
KINDERGARTEN SAMPLES COMBINED

Subscale, Item No., and Item Factor 1 Factor 2 Mean SD

Cognitive competence:
1. Good at puzzles .................... .39 3.2 .77
5. Gets stars on paper ................. .37 3.1 .95
9. Knows names of colors .............. .57 3.6 .60
13. Good at counting .................. .43 3.6 .61
17. Knows alphabet ................... .48 3.6 .67
21. Knows first letter of name ........... .58 -.33 3.6 .62
Physical competence:
3. Good at swinging .................. .19 3.6 .84
7. Good at climbing .................. .33 3.4 .77
11. Can tie shoes ..................... .42 2.8 1.12
15. Good at skipping ................ .34 3.4 .84
19. Good at running .................... .23 3.4 .76
23. Good at hopping ................... .22 .30 3.4 .75
Peer acceptance:
2. Has lots of friends .................. .36 3.2 .79
6. Stays overnight at friends'. .......... .47 3.1 .92
10. Has friends to play games with ...... .23 3.1 .86
14. Has friends on the playground ....... .36 3.2 .79
18. Gets asked to play with others....... .44 3.1 .81
22. Eats dinner at friends' house ........ .61 2.7 1.01
Maternal acceptance:
4. Mom smiles .................. .... .52 3.3 .67
8. Mom takes you places you like ...... .52 3.1 .80
12. Mom cooks favorite foods ........... .53 3.0 .75
16. Mom reads to you ................ .61 3.0 .96
20. Mom plays with you ................ .70 2.5 1.04
24. Mom talks to you ................... .62 3.1 .91
NOTE.-N = 145.

It should be noted that, since the item

scales; it is considered to reflect perceptions
of social acceptance. means for the competence subscales, in par-
Item Means a'nd Standard Deviations ticular, were skewed toward the upper end of
the scale, the range of scores was restricted,
As can be seen in Tables 2 and 3, the
which in turn attenuated the magnitude of
majority of means are in the range of 3.0-3.6,
these reliability estimates. That is, the over-
indicating that young children tend to report
whelming majority of children's item scores
relatively positive feelings of competence and
were either 3 or 4. Paradoxically, therefore,
acceptance. Standard deviations indicate that
although children responded consistently to
there is still considerable variability, even
these items in terms of scores at the upper end
though judgments are being made in the up-
of the scale, the restricted range necessarily
per ranges of the scale. The use of the upper
leads to lower statistical estimates of reliability.
ranges is not thought to reflect social desirabil-
ity response tendencies so much as the Subscale
young Means and Standard Deviations
child's blurring of the boundaries between There-
subscale means and standard devia-
ality and the wish to be competent tions
or are
ac-presented in Table 5. These subscale
cepted, as anticipated. means, like the item means, are skewed in the
direction of positive judgments, reflecting the
Reliability
tendency for young children to report rela-
Subscale reliabilities, presented in Table
tively positive feelings of competence and so-
4, were assessed by employing coefficient oa cial acceptance. Scores are somewhat higher
that provides an index of internal consistency.
for the two competence subscales, compared
If one looks at individual subscales, it can be
to the two social acceptance subscales. Con-
seen that these values range from .50 to .85.
sistent with this pattern, the standard devia-
When one combines subscales according to
tions are somewhat lower for the competence
their designated factors, these reliabilities in-
than the social acceptance subscales.
crease substantially, falling within a range of
.75-.89. The reliability of the total scale, all 24 A 4 x 4 (group x subscale) analysis of
items, is in the mid- to high .80s. variance, with subscale as a repeated mea-

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
 Harter and Pike 1975

TABLE 3

FACTOR PATTERN AND ITEM MEANS AND STANDARD DEVIATIONS FOR THE FIRST- AND
SECOND-GRADE SAMPLES COMBINED

Subscale, Item No., and Item Factor 1 Factor 2 Mean SD

Cognitive competence:
1. Good at numbers ................ .51 3.2 .73
5. Knows a lot in school ............. .63 3.5 .64
9. Can read alone .................. .50 3.4 .73
15. Can write words ................. .65 3.6 .58
17. Good at spelling ................. .51 3.4 .65
21. Good at adding .................. .40 3.5 .62
Physical competence:
3. Good at swinging ................ .22 3.7 .60
7. Good at climbing ................ .48 3.4 .80
11. Good at bouncing ball............ .43 3.5 .71
15. Good at skipping................. .33 3.7 .63
19. Good at running ................ .50 3.4 .70
23. Good at jumping rope ............ .40 3.1 1.02
Peer acceptance:
2. Has lots of friends ............... .67 3.1 .85
6. Others share their toys ........... .27 3.3 .78
10. Has friends to play games with ... .60 3.0 .90
14. Has friends on the playground .... .67 3.2 .89
18. Gets asked to play with others .... .72 3.1 .85
22. Others sit next to you ............ .67 3.1 .81
Maternal acceptance:
4. Mom lets you eat at friends' ...... .44 2.8 .89
8. Mom takes you places you like ... .58 3.1 .95
12. Mom cooks favorite foods ......... .63 3.1 .77
16. Mom reads to you................ .61 2.7 1.13
20. Mom lets you stay overnight...... .51 2.9 1.01
24. Mom talks to you ................ .50 3.0 .94
NOTE.-N = 104.

sure, revealed a significant effect for

calsubscale,
competence and the two social accepta
F(3,693) = 92.02, p < .001. As can be seen infor the second graders, howe
subscales
Table 5, this primarily indicates thatdid scores
not fit this pattern.
were significantly higher for the two compe-
Correlations
tence subscales, where most means were at orbetween Child and Teacher
Ratings
close to 3.4, than for the two social acceptance
The intercorrelations between child and
subscales where the means ranged from 2.8 to
3.1. While there was no group effect,teacher judgments were calculated across all
there
was a group x subscale interaction subjects
that since
ap- differences between age groups
were
pears primarily to be because, for the small.
first and These values were .37 (p < .001)
forscores
second graders only, peer acceptance cognitive competence, .30 (p < .005) for
physical
were higher than maternal acceptance competence, and .06 for social accep-
scores,
which was not the case for the tance. While these correlations are moder-
younger
groups, F(9,693) = 5.85, p < .01. ately weak, the pattern indicates that agree-
ment between pupil and teacher is highest in
Intercorrelations among Subscales the cognitive competence domain, next high-
Table 6 presents the intercorrelations
est in the physical domain, and virtually negli-
among the four subscales for each of thefor
gible four
peer acceptance.
groups. The clearest patterns obtained, consis-
Validity
tent with the factor analysis, are for Data
the two
Convergent validity.-As one index of
competence scales to intercorrelate moder-
ately for each group and for the two social of children's judgments, we con-
the validity
acceptance subscales to intercorrelate some-
ducted an inquiry after the measure had been
what more highly. Among preschoolers,administered
kin- asking children the bases for
dergartners, and first graders, peer their responses in the two competence do-
acceptance
also correlates moderately with bothmains. Children were asked, "How do you
cognitive
and physical competence, as does know maternal
you are good at/not good at [depending
acceptance. The correlations between on physi-
the child's initial response] this [activity

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
 [ e

1J

Se - rt c
z oO C~CoO fx 4
?

SC

[-,

C/2

V
U
0

U
"0
z
o / )"a
C1 1
EzJ 4 U
0z

a, r r"0..~c
*3 * V.

p;a
z
,<

U "0

o cc b"

?r o Wu0 V~ a" c~

ZL~Z

.e . (

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
 Harter and Pike 1977

TABLE 5

SUBSCALE MEANS AND STANDARD DEVIATIONS FOR EACH GROUP

COMPETENCE SUBSCALES ACCEPTANCE SUBSCALES

Cognitive Physical Peer Maternal

Mean SD Mean SD Mean SD Mean SD

Preschool (N = 90) ........ 3.4 .45 3.2 .49 3.0 .56 3.1 .59
Kindergarten (N = 56) ..... 3.6 .41 3.4 .35 2.9 ,56 2,9 ,58
Preschool and
kindergarten combined .. 3.5 ,43 3.3 .46 2.9 .56 3.0 .59
First grade (N = 65) ....... 3.4 .37 3.4 .38 3.1 .55 2.8 .60
Second grade (N = 44) .... 3.5 .31 3.4 .40 3.1 .55 2.8 .56
First and second
grades combined ...... 3.4 .35 3.4 .39 3.1 .55 2.8 .58

TABLE 6

INTERCORRELATIONS AMONG SUBSCALES FOR EACH GROUP

Cognitive Physical Peer

Competence Competence Acceptance

Physical competence:
Preschool .............. .56***
Kindergarten ........... .43***
First grade ............. .55***
Second grade............ .43**
Peer acceptance:
Preschool ............... .56*** .48***
Kindergarten ........ .... .45*** .42***
First grade ............. . 59*** ,50***
Second grade........... .32* 08
Maternal acceptance:
Preschool ...........,. .48*** .43*** .64***
Kindergarten ........... .27* .50*** .62***
First grade ............. .51"** ,48*** .66***
Second grade ........... .32* .00 .80***

* p < .025,
** p < .01.
*** p < .001.

specified]? How can you tell?" The purpose
of this procedure was to determine (a)
whether children could give reasons, and if
so, (b) whether they were compelling in the
sense that they bolstered or supported the par-
ticular self-judgment they had given previ-
ously.
Systematic data (Chao, Harter, Adams, &
Strop, Note 3) were available for a sample of 43
first graders and 48 second graders who were
asked about three cognitive skills (reading,
spelling, and writing) and two physical skills
(climbing and running). For the cognitive
skills, 96% of the children readily gave
specific reasons for why they felt that they
were competent or not competent. These fell
into the following five categories for which
sample responses are provided: (a) perfor-
mance feedback (19%): "I get the hardest
words right on a test," "the teacher tells me";
(b) possess/perform component skills (11%):
"I sound out the letters," "I memorize the
words," "I draw straight letters"; (c) specific
demonstrations (32%): "I started reading
when I was 3," "I can write words like 'cat'
and 'dog,' " "I can read two whole books," "I
can write in handwriting," "I don't have to
read out loud, I can think it up in my mind";
(d) routes to developing skills (20%): "I prac-
tice a lot," "I can spell 'cause I read a lot," "I
practice on my flash cards," "My mom and
dad helped me learn how"; and (e) habitual
activity (14%): "I read a lot at home," "I do
writing every day," "I've spelled a lot before."
In the physical domain, the categories
differed somewhat from the cognitive domain,
and there were also within-domain differ-
ences in the percentage of responses for

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
1978 Child Development

climbing and running. The categories, includ-significantly lower than the scores (mean =
ing sample responses, were as follows: (a) 3.3)
so- of those who were promoted, t(22) = 3.5,
cial comparison (51% for running, 2% pfor < .005.
climbing): "I got to race a lot and win," "I wasFor the social domain, we examined the
first place in running in gym," "The boys say
perceived peer acceptance scores for kinder-
they can run faster but at races I can beat
gartners and first and second graders who re-
them," "I'm the best in climbing races"; (b)
cently moved and who had attended this par-
habitual activity (18% for running; 31% for ticular school for less than 2 months. We
climbing): "I run a lot," "I practice jogging,"
hypothesized that these children would have
"I do a lot of feet work in the gym," "I climb a
lower peer acceptance scores than children
lot on bars," "I've been climbing for a longwho had been in the school for a minimum of
time"; (c) specific demonstrations (26% and
1 year. The scores of the 10 "new" children
52%, respectively): "I can run around were the significantly lower (mean = 2.9) than a
block a couple of times," "I run a lot in foot-
comparison group of children, matched for
ball and tag 'em down," "I climb up to my
age and gender, whose scores (mean = 3.3)
treehouse," "I can do chin-ups"; (d) some- indicated greater peer acceptance, t(18) = 2.7,
body teaches (1% and 40%, respectively):
p< .01.
"Mom taught me," "Somebody taught me to
climb the jungle gym"; and (e) don't injureIn the physical domain, we have exam-
self (4% and 3%, respectively): "I hardly ever
ined the validity issue with regard to the
trip and fall," "I don't fall off and scratch my-
scores of children who were preterm infants.
self." (A total of 96% of the responses about
Prematurity is frequently associated with de-
running and 97% of the responses about velopmental lags in gross motor skills. From
climbing could be coded in these categories.)
preschools that had participated in our stud-
ies, we were able to obtain information from
Therefore, for both the cognitive and teachers as to which of the children they were
physical domains, the findings demonstratecertain had been born preterm. These were
that children can provide very definite rea-
compared with a sample of preschoolers who
sons for their alleged competencies; more-
were known to have been full-term infants.
over, they volunteer these readily. Although
Group differences in teacher ratings for the
we do not have systematic data for the youn-
physical domain were considerably lower for
ger ages, many of these children spontane-
the preterm group (mean = 2.3) compared
ously elaborated on (and sometimes demon-with the full-term group (mean = 3.1), t(14) =
strated) their prowess during the course of the
3.4, p < .005. Correspondingly, the physical
normal administration, and these comments competence scores of the eight children who
reveal that they too have specific reasons had
for been preterm infants were found to be
their judgments. Furthermore, although lower
the (mean = 2.8) than the scores (mean =
sample responses presented were those 3.3) of-of children who had been full-term in-
fered by children judging themselves to be fants, t(14) = 2.9, p < .01.
competent, children rating themselves as in-
competent also gave plausible responses (e.g., We have begun to examine the validity of
"I can't spell words on tests," "I draw crooked
the maternal acceptance subscale in one study
letters," "I watch too much TV," "I can't doof childhood depression (Harter & Wright,
Note 4). Our prediction was that depression in
twirls on the jungle gym," "I'm the last when
we run"). Therefore, the overall pattern is young
one children (defined in terms of dysphoric
of convergence between the initial perceivedmood and lack of energy or interest) would be
competence judgments and the reasons chil- directly related to lack of maternal acceptance.
dren offered for these perceptions. In this study, we did not have a group of se-
verely depressed children. However, within
Discriminant validity.-As one test the of normal range of scores for kindergartners
validity in the cognitive domain, we made andthe first and second graders, we found the
prediction that children held back in first
correlation between our depression/cheer-
grade for academic reasons should score lower
fulness measure and maternal acceptance to
on the cognitive competence subscale than be .48, p< .001.
those who were promoted to the second
grade. Over a 2-year period we identified 12 Finally, although our new paternal accep-
children who had been held back, and we tance scale has not yet been integrated into
compared these children with a sample ofthe12 versions reported on in this paper, an in-
children, matched on age and gender, fromteresting study has just been completed on
the pool of those who had been promoted. young children with abusive fathers (Kelty,
The cognitive competence scores of those Note 5). Examining 11 such children, it was
held back (mean = 2.4) were found to found
be that their fathers' acceptance scores

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
 Harter and Pike 1979

quartiles
(mean = 2.6) were significantly lower than (3.6 and 2.3, respectively, for the
the scores (mean = 3.2) of a groupcognitive
of 13 domain), differences in the per-
ceived cognitive competence of these two
nonabused children from the same preschool,
t(22) = 3.4, p < .005. subgroups are highly significant (3.8 vs. 2.6),
t(102) = 5.9, p < .001. Thus, for children who
Predictive validity.-In one study, Bierer
fall at either end of the competence con-
(1981) examined the relationship between
tinuum, there is much more convergence be-
first and second graders' perceived cognitive
tween teacher and child ratings than for those
competence and their preference for chal-
falling within the mid-ranges of the distribu-
lenge on a behavioral task involvingtion.
subjects'
choice of puzzles, varying in difficulty level.
It was initially hypothesized that per- Discussion
ceived cognitive competence would predict
difficulty-level preferences. This correlation The attempt to devise a pictorial self-
(r = .42) was significant, p < .005. However, report measure of young children's percep-
further examination indicated that it was at- tions of their competence and social accep-
tenuated because a subgroup of children was tance would appear to have been successful.
present whose perceptions of their compe- Children eagerly respond to the pictorial for-
tence appeared to be inflated-that is, whose mat, they comprehend the items, and the
scores were at least 1.2 higher (on a four-point psychometric properties of the scale seem
scale) than their teacher's ratings of their cog- sound. The item scores and standard devia-
nitive competence on the same items. This tions revealed reasonable variability, indicat-
subgroup tended to select puzzles that were ing that the scale is sensitive to individual
much easier than one would expect, based on differences in perceived competence and ac-
their perceived competence, although theirceptance among young children.
choices were consistent with their actual
The reliability, as assessed through in-
competence, as judged by the teacher.dexes
In of internal consistency, was found to be
terms of the validity question, these findings
acceptable. Several forms of validity were also
revealed ti.-+ for pupils whose ratingsexamined.
are In normative samples, the reasons
either congruent or lower than the teachers,
children gave for their self-perceptions were
their perceived cognitive competence is pre-consistent with their judgments on the items
dictive of their actual behavior. That is, these
themselves and were quite plausible. This
perceptions appear to mediate their behav-
suggests that the ratings are valid, in the sense
ioral preference for challenge. However, the
that young children's self-perceptions of their
presence of overraters in the sample competencies
at- appear to be based on specific
tenuates the predictive validity of this sub-
behavioral referents.
scale.

Correlations between child and teacher The findings also indicated that scores on
the various subscales do discriminate be-
ratings.-In the introduction, it was sug-
tween groups of children predicted to differ in
gested that the tendency of young children to
be somewhat inaccurate observers of their each domain. For example, children new to a
own competencies does not necessarily indictschool setting reported lower peer acceptance
than those who have attended the school for a
the validity of the instrument. The findings
presented indicate that the correlations be-year or more. Children who have been held
tween self- and teacher ratings in the two back a grade for academic reasons reported
lower perceived cognitive competence than
competence domains are significant, although
those experiencing normal promotion. Chil-
they are moderately weak, at best, consistent
with our expectation. Nevertheless, we did dren who were preterm infants, with related
delays in motor development, had lower
find that, for the competence domains, teacher
and child ratings were more highly correlatedphysical competence scores than children
within the same domain (cognitive = .37, who had been born full-term. Thus, the vari-
ous subscales would appear to discriminate
physical = .30) than they were across the two
domains (teacher-cognitive/pupil-physical = clearly between a given subgroup for whom
there is reason to expect relatively low scores
.11; teacher-physical/pupil-cognitive = .16).
and children, matched for age and gender,
Thus, while we have not relied heavily on this
from
type of external validity, the pattern suggests the normative sample. In addition, chil-
that children's competence judgments are re- dren judged by teachers to be very competent
lated to their actual competence. scored considerably higher than those whom
teachers judged to be low in competence.
Moreover, when one examines the per-
ceived competence scores of children whom At a more theoretical level, the factor pat-
tern obtained with this instrument is of inter-
the teachers rate as in the top and bottom

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
1980 Child Development
more critical of others than of the self (Gesell
est, since it provides certain clues with regard
to the structure of the young child's self-
& Ilg, 1946; Stipek, 1981).
descriptions. The two-factor solution revealed
With regard to competence judgments,
one factor comprised of the cognitivethe and
verbal interview data revealed that self-
physical competence subscales and a second
perceptions of skill were directly tied to
factor comprised of the peer and maternal ac-
ceptance subscales. We have labeled these
specific behaviors emitted by the self. We
have yet to inquire about the bases for judg-
factors "general competence" and "social ac-
ments concerning social acceptance. How-
ceptance." The single competence factor im-
ever, we have collected interview data on the
plies that young children do not make a clear
distinction between what we identified as perceived routes to both competence and so-
cial acceptance. We have systematically asked
cognitive and physical domains. Competence
children what the child depicted as incompe-
at one type of skill is associated with compe-
tent or socially unaccepted would have to do
tence at the other. One is either "good at do-
to become like the child pictured as compe-
ing things" or one is not. These skill domains,
tent or accepted. The responses to compe-
however, are distinguished from social accep-
tence involve self-improvement, primarily
tance by peers and by mother.
through instruction (I learn from the teacher),
The structure of young children's self- or personal effort (I practice a lot, try harder,
perceptions across these domains is less dif-etc.), with a gradual shift toward high levels of
ferentiated than the structure obtained for personal effort over the age span of 4 to 7
older children (Harter, 1982) where we find (Chao, Harter, Adams, & Strop, Note 3).
that cognitive and physical skills clearly In contrast, the routes to peer acceptance
define separate factors.2 This developmental
involve behaviors designed to influence
difference is consistent with findings indicat-
others. Here we find a gradual shift from rela-
ing that the structure of the self becomes more
tively naive solutions (find friends, just ask
differentiated with age (see Harter, Note 1). In
people to be your friends) to social strategies
addition, the self-structure would appear to be
such as being nice, helpful, polite, and kind.
more highly related to mental age than The spontaneous mention of these strategies
chronological age. In one study we employedincreases from a low of 18% among 4-5-year-
retarded children whose mental ages ranged olds to 46% in first graders and 65% in second
from 5 to approximately 8, the same range one
graders.
would expect for the young normal-IQ chil-
dren in the present study. Consistent with the These findings are interesting in light of
findings reported here, the retarded pupils the recent emphasis on social skills (see
did not make a distinction between the cogni-Asher, Oden, & Gottman, 1977; Bash &
tive and physical competence domains, Camp, 1980; Gottman, Gonso, & Rasmussen,
whereas social acceptance defined a separate 1975; Hartup, 1979, 1983; Ladd & Oden,
factor (Silon & Harter, in press). 1979; Spivack & Shure, 1974). Across these
studies it has been revealed that a variety of
Inspection of the subscale means indi- social skills are associated with peer popular-
cates a general tendency for scores to be ity and social acceptance. Moreover, they re-
skewed toward positive self-evaluations, al- veal that elementary school children possess
though this tendency was greater for the two an awareness of how social skills influence
competence subscales than for the two social
their acceptance by peers. Our own findings
acceptance subscales. This pattern appears suggest that, at the youngest ages, children
plausible since judgments about one's com- have not yet acquired the knowledge concern-
petencies may be more intimately related to ing this relationship in the social domain, al-
one's appraisal of the self, in contrast to judg-
though they do seem to appreciate the need
ments about social acceptance, which may be for skill development in the cognitive and
influenced by one's view of the characteristics
physical domains. Gradually, over the early
of these particular others. Since fantasies grades, they come to appreciate the need to
about the ideal self intrude upon judgments of employ social strategies in order to obtain
the real self at this age level, the competence friends.
scores are likely to be somewhat inflated. To
the extent that social acceptance items pull for This pattern raises numerous questions
judgments of others, these scores would be for further study. For example, in what ways
expected to be lower since findings have dem- are social skills different from those in the cog-
onstrated that young children are likely to be nitive and physical domains? Why should the
2 In the revision of the original perceived competence scale, we have determined that children
age 8 and older make distinctions among five domains (scholastic, athletic, appearance, social accep-
tance, and conduct), as revealed by a clean five-factor solution (Harter, 1983).

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
 Harter and Pike 1981

understanding of these two types ofand interest

skills pur- for age-appropriate activities. To
the extent
sue a different developmental course? In ad- that the child's mood and energy
level may beof
dition, how do the perceived characteristics critical mediators of behaviors
others influence one's judgments ofleading
social to ac-
the development of new skills, it
would be important
ceptance? In light of these considerations, we to assess these predictors
would urge that our instrument not as be
early as possible. Furthermore, a domain-
treated
specific
solely as a measure of "self-concept" sincemeasure
the allows one to determine
social acceptance construct tapped may
which in- best predict the mediators and
domains
behaviors be-
volve a number of dimensions extending of interest.
yond one's perceptions of the self per se.
Second, the degree to which a young
Differences between the two child's
compe-judgments are inaccurate might be im-
tence domains and the social domainportant to examine. Normative or age-
of peer
appropriate
acceptance were also revealed in the examina- distortions may not be cause for
tion of the accuracy of children's judgments.concern. However, it may be important to de-
The correlations between teacher and child tect either extreme inflation of one's abilities
ratings for the two competence domains were or the unrealistic portrayal of oneself as ex-
significant, though moderately weak; how- tremely incompetent. Furthermore, a child
ever, they were negligible in the social do- may show these inaccuracies in some domains
main. The greater congruence in the compe- more than others, and these particular distor-
tence domain may be because of clearer tions may well have behavioral correlates.
sources of information on which to base one's Findings with somewhat older children indi-
judgments. Our findings suggest that, in the cate that, by third and fourth grades, there are
cognitive domain, performance feedback is behavioral patterns associated with extreme
beginning to emerge as a criterion for perfor- tendencies to overrate or underrate one's cog-
mance, whereas in the physical domain, social nitive competence. For example, both of these
comparison is becoming the basis for judg- inaccurate subgroups tend to avoid behavioral
ments of competence. Of particular interest is preference for challenge compared with those
the finding that social comparison is used children who accurately rate their compe-
more frequently for the activity of running, tence (Bierer, 1981).
which appears to be a more competitive activ- Third, there would appear to be a need
ity, than for climbing. for an instrument to assess the self-
Our data are consistent with Ruble and perceptions among special subgroups of chil-
Frey's (Note 6) findings that, in the domain of dren who may be under particular types of
academic achievement, social comparison is stress. Children of divorce, of abusive parents
not consistently employed in the early grades. and with learning disabilities or physica
Our findings also indicate that social compari- handicaps are all special groups that have
son effects may be somewhat domain-specific come to the recent attention of basic research-
since social comparison does form the basis ers, clinicians, and those engaged in socia
for judgments of certain physical skills. We policy. However, as has been pointed out, no
have yet to examine the bases on which chil- all children necessarily suffer from events that
dren make judgments in the social domain. have been categorically identified as stressful
However, the lack of congruence between (Garmezy & Rutter, 1983). A variety of indi
child and teacher ratings of social acceptance vidual difference variables, including self
may result from several factors: performance concept, have been implicated as factors in-
feedback may be less salient, children may be fluencing the child's ability to withstand stress
less able to employ social comparison in this and cope adaptively. Thus, a domain-specific
domain, and/or children and adults may em- measure might well be useful in predicting
ploy different criteria. Further research in this children's reactions with an eye toward deter
area would be fruitful. mining which type of profile is associated
with resiliency and adaptation, or its
Finally, to what uses might such an in- counterpart. In conclusion, although there are
strument be put, particularly given the several theoretical issues requiring further re-
qualification that young children's judgments search, we believe that there are a number of
are not very accurate? First, among normative
uses to which this instrument might well be
samples, scores may be useful in predicting
put in order to illuminate our understanding
behaviors, motivations, and/or emotional reac-
of the young child.
tions of interest. Our own findings (Harter &
Wright, Note 4) indicate that the social accep- Reference Notes
tance subscales, particularly the maternal sub-
scale, are significantly correlated with the 1. Harter, S. Supplementary description of the
child's self-reported mood as well as energy Self-Perception Profile for Children: Revision

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms
1982 Child Development

of the Perceived Competence Scale for Chil- in children. Child Development, 1975, 46, 709-
dren. Unpublished manuscript, University of718.
Denver, 1983. Harter, S. The perceived competence scale for chil-
2. Harter, S., & Pike, R. Procedural manual to ac-dren. Child Development, 1982, 53, 87-97.
Harter, S. Developmental perspectives on the self-
company the Pictorial Scale of Perceived Com-
petence and Social Acceptance for Young Chil- system. In E. M. Hetherington (Ed.), P. H.
dren. Unpublished manuscript, University ofMussen (Series Ed.), Handbook of child psy-
Denver, 1983. chology. (Vol. 4): Socialization, personality,
3. and social development. New York: Wiley,
Chao, C., Harter, S., Adams, P., & Strop, J. Di-
mensions underlying children's perceptions of1983.
competence. Unpublished data, University Harter,
of S. Competence as a dimension of self-
Denver, 1983. evaluation: Toward a comprehensive model of
4. Harter, S., & Wright, K. The relationship be- self-worth. In R. Leahy (Ed.), The development
tween social support, depression, and young of the self. New York: Academic Press, in
children's perceptions of competence and so-press.
Hartup, W. W. Peer relations and the growth of so-
cial acceptance. Unpublished manuscript, Uni-
versity of Denver, 1984. cial competence. In M. W. Kent & J. E. Rolf
5. Kelty, E. Perceived competence and social ac-(Eds.), The primary prevention of
ceptance in young children of abusive and non-psychopathology. (Vol. 3): Promoting social
abusive parents. Unpublished manuscript, competence and coping in children. Hanover,
University of Colorado, 1983. N.H.: University Press of New England, 1979.
6. Ruble, D. N., & Frey, K. S. Self-evaluation Hartup,
and W. W. Peer relations. In E. M. Hethering-
social comparison in children: Developmentalton (Ed.), P. H. Mussen (Series Ed.), Handbook
changes in knowledge and function. Paper pre-of child psychology. (Vol. 4): Socialization, per-
sented at the biennial meeting of the Societysonality, and social development. New York:
for Research in Child Development, Detroit,Wiley, 1983.
1983. Ladd, G. W., & Oden, S. The relationship between
peer acceptance and children's ideas about
helpfulness. Child Development, 1979, 50,
402-408.
References
Piers, E., & Harris, D. The Piers-Harris Children's
Asher, S.R., Oden, S. L., & Gottman, J. M. Chil-Self-Concept Scale. Nashville, Tenn.: Coun-
selor Recordings and Tests, 1969.
dren's friendships in school setting. In L. G.
Ruble, D. N. The development of social comparison
Katz (Ed.), Current topics in early childhood
education (Vol. 1). Norwood, N.J.: Ablex, 1977.processes and their role in achievement-related
Bash, M. A. S., & Camp, B. W. Teacher training self-socialization.
in In E. T. Higgins, D. N.
the Think Aloud classroom program. In G. Ruble, & W. W. Hartup (Eds.), Social cognition
Cartledge & J. F. Milburn (Eds.), Teaching so-and social development: New York: Cambridge
cial skills to children. New York: Pergamon, University Press, 1983.
1980. Silon, E., & Harter, S. The assessment of perceived
Bierer, B. Preference for challenge among children competence, motivational orientation, and anx-
who over-rate, under-rate, and accurately rate iety in segregated and mainstreamed educable
their cognitive competence. Unpublishedmentally retarded children. Journal of Educa-
Ph.D. dissertation, University of Denver, 1981.tional Psychology, in press.
Spivack, G., & Shure, M. B. Social adjustment of
Coopersmith, S. The antecedents of self-esteem.
San Francisco: W. H. Freeman, 1967. young children. San Francisco: Jossey-Bass,
1974.
Garmezy, N. & Rutter, M. (Eds.). Stress, coping,
and development in children. New York: Stipek, D. J. Children's perceptions of their own
McGraw-Hill, 1983. and their classmates' ability. Journal of Educa-
tinal Psychology, 1981, 73, 404-410.
Gesell, A., & Ilg, F. The child from five to ten. New
York: Harper & Row, 1946. Wylie, R. Theory and research on selected topics.
Gottman, J. M., Gonso, J., & Rasmussen, B. Social (Vol. 2): The self concept. Lincoln: University
of Nebraska Press, 1979.
interaction, social competence, and friendship

This content downloaded from

154.59.124.141 on Sun, 15 Jan 2023 02:38:22 UTC
All use subject to https://about.jstor.org/terms