Printer Friendly

An investigation of self-concept in gifted children.

ABSTRACT: The study explored several issues respecting the self-concept in children categorized as gifted: (a) the relative independence of specific components of self-perceptions; (b) the way in which these factors relate to global self-esteem; and (c) the extent to which a developmental process operates in the evolution of the self-concept. Data were collected from a sample of pupils enrolled in enrichment classes, grades 5 through 8. Measures included the Self-Perception Profile for Children and a teacher-rating measure of pupil attributes. The results confirmed the relative independence of the specific components and showed that self-perceptions of social and scholastic competence and of physical appearance were the major contributors to the self-concept. There was no evidence, however, for the operation of a developmental process.

There has been a considerable expansion in theoretical and empirical efforts to analyze children's self-perceptions over the past 10 years or so (cf. Byrne, 1984; Harter, 1983; Shavelson, Hubner, & Stanton, 1976). One of the more interesting of the recent theoretical developments is that offered by Harter (1983, 1986). Harter's work has been especially useful in two ways. First, she has treated self-concept as a multidimensional construct rather than a unitary entity; second, she has provided measuring instruments appropriate for assessing aspects of her model.

The present study is an effort to assess certain postulates of her model with a sample of pupils categorized as gifted. Though the question of self-concept in the gifted has received considerable research attention cf Schneider, 1987), most efforts have entailed comparisons of gifted and nongifted samples in terms of global selfconcept. Relatively little attention has been paid to the structure of the self-concept in gifted children, the issue of concern in the present study.

Two key postulates of that model are examined here. First, Harter hypothesized that selfconcept in children is composed of a global component and a set of specific components (having to do, for example, with perceived social competence and academic competence). She further hypothesized that the global component of the selfconcept represents a relatively independent construct, rather than an additive product of the specific components. A similar formulation has been advanced by Byrne (1984), Shavelson et al. (1976), and Winne and Marx (1981).

A second key postulate of the model is that a developmental process operates in the case of the self-concept in the sense that the specific components of self-esteem become more differentiated over the life span. Thus, whereas very young children might show a great deal of overlap among social, academic, and athletic competence, these areas of the self-concept will be more independent in older children. It follows, as well, that the way in which the specific components relate to global self-worth will change with the age of the child.

Some data bearing on the two postulates have been reported in studies using an earlier self-report measure, the Perceived Competence Scale for Children (Harter, 1982). This measure is scored in terms of four factor scores: Cognitive Competence, Social Competence, Physical Competence, and General Self-Worth.

Both Harter (1982) and Byrne and Schneider (1988) have reported correlational and factor analytic results supporting the postulate that the specific aspects of self-concept exist relatively independently of one another and of the global self-worth construct. The latter researchers demonstrated these findings for samples of pupils from both regular and gifted classrooms.

These researchers obtained somewhat contradictory findings respecting the way in which the specific dimensions of perceived competence related to global self-worth. Thus, Harter (1982) reported that the physical and social competence factors were most closely related to global self-worth for her sample of regular classroom pupils, while Byrne and Schneider (1988) found that the strongest links with global self-worth were obtained with the cognitive and social factors, and this pattern was true for both the gifted and regular samples of pupils. This is a particularly interesting issue so far as gifted pupils are concerned; we might expect on intuitive grounds that cognitive or academic aspects of the self-concept would receive relatively high weights in the determination of global self-worth because these children have likely received considerable rewards for academic achievement.

The Harter (1982) and Byrne and Schneider (1988) studies also contain support for the hypothesis that the specific components of self-esteem become more differentiated with age. In both cases, correlations among the specific self-concept factors became progressively weaker across age groups. Though Byrne and Schnelder (1988) included both regular and gifted pupils in their samples, they did not provide separate analyses for those groups. The present study provided an opportunity to explore this developmental hypothesis with a sample of gifted children. This, too, is a particularly interesting question in the case of gifted children. Harter (1983, 1986) has suggested that the developmental process is associated with intellectual maturity; hence, it might be expected that these children will show a differentiation among specific aspects of the self-concept relatively early.

The current study was designed, then, to explore issues relating to (a) the relative independence of the specific components of the self-concept, (b) the components of global self-worth, and (c) the developmental process whereby the components become more differentiated with age. The study extends earlier work in this area by focusing on a sample of gifted children and by employing a recent revision of the Perceived Competence Scale for Children. This new instrument, the Self-Perception Profile for Children (Harter, 1985), is based on reliability and validity studies done in connection with the earlier instrument; and it provides a more differentiated analysis of self-concept through six factor scores: Scholastic Competence, Social Acceptance, Athletic Competence, Physical Appearance, Behavioral Conduct, and Global Self-Worth.

We were also able to explore issues relating to gender differences in the self-concepts of gifted children. Only two other studies have explored this issue, and they yielded mixed results. Thus, Milgram and Milgram (1976) reported higher global self-concept scores for gifted girls compared with gifted boys. Schneider, Clegg, Byrne, Ledingham, and Crombie (1989), on the other hand, reported higher global self-worth and higher perceived physical competence scores for boys from samples of 8th- and 10th-grade gifted pupils (though not 5th grade). The present study provided an opportunity to further explore this issue with the more psychometrically refined measure, the Self-Perception Profile for Children.



The sample was composed of 280 pupils drawn from four grade levels, 5 through 8. All were enrolled in self-contained enrichment classes of an urban school board. The majority of the pupils had been selected for this gifted program while in fourth grade, with the selection procedure based on group and individual IQ test performance, standardized achievement test scores, and teacher ratings. A 90th percentile cut-off was used in the case of the standardized tests. Further evidence that this represented a highly select group may be found in the standardized achievement test scores administered while the students were enrolled in the enrichment classes: average percentile scores for the pupils ranged from 88 to 93 (based on citywide norms).

Measures and Procedures

The Self-Perception Profile for Children (SPPC).

This self-report measure is designed to assess self-perceptions of competence. It was developed by Harter (1985) from her earlier instrument, the Perceived Competence Scale for Children (Harter, 1982). The scale was administered to the pupils in the classroom setting by research personnel.

The SPPC is composed of six subscales, each containing six items. Each item

entails a two-step process whereby the child first indicates which of two alternatives best describes them (e.g., "some kids are happy with the way they look" vs.

other kids are not happy with the way they look") and, second, indicates the extent to which the description is true of them. This format is designed to reduce the operation of socially desirable responding.

The six subscales include the general self-concept construct, Global Self-Worth, and five specific aspects of the self-concept: Scholastic Competence, Social Acceptance, Athletic Competence, Physical Appearance, and Behavioral Conduct. These subscores are based on factor analytic analyses (Harter, 1982, 1985). Four of these subscales are virtually identical with subscales contained in the Perceived Competence Scale for Children; these are the subscales Scholastic Competence, Social Acceptance, Athletic Competence, and Global Self-Worth. The psychometric properties of these subscales have been well established (Byrne & Schneider, 1988; Harter, 1982). Additional psychometric data for the larger SPPC instrument have also been presented in the manual Harter, 1985). These are based on internal consistency analyses of the subscales and additional factor analytic efforts.

Scale for Rating Behavoral Characteristics of Superior-Students (SRBCSS). This teacher-rating measure was developed by Renzulli, Hartman, and Callahan (1971) for assessing the potential of children for gifted programs. It was used in the present study as a source of information about the cognitive, social, and academic competencies of the pupils and was included in analyses of the determinants of global self-worth. Rating forms were prepared for each pupil in the class and teachers were asked to complete the ratings within a 2-week period.

The instrument is composed of 37 items divided among four subscales; these subscales are labeled Learning, Motivation, Creativity, and Leadership. A four-point rating scale is provided for each item. The following is an example of an item from the Learning subscale: "Is a keen and alert observer; usually 'sees more' or gets more' out of a story, film, etc. than others.

Support for the internal consistency and test-retest reliability of the four subscales has been reported by Renzulli et al. (1971). Some positive validity results have also been reported by Renzulli et al. and Ashman and Vukelich (1983).


Descriptive Data

Table 1 presents mean SPPC scores by gender and grade level. Analyses of variance revealed no significant effects for grade level or for the interaction of grade level and gender. There were, however, significant gender effects for the Physical Appearance, F(1,224) = 11. 16, p < .001, Behavior Conduct, F(1,224) = 5.06, p < .05, and Athletic Competence, F(1,253) = 21.44,p <.001, subscores.

These gender differences are shown in the first two columns of Table 2. It can be seen that boys were displaying higher scores on Appearance and Athletic Competence; and girls, higher scores on Conduct. These results are generally consistent with those of Schneider et al. (1989) and Harter (1985). Results relating to the Global Self-Worth factor are also of interest. Harter (1985) reported higher global scores for boys than for girls for pupils in regular fifth- through eighth-grade classrooms, whereas Schneider et al. (1989) reported higher scores for males in eighth-grade regular and enrichment classrooms. Our results show no significant differences between boys and girls on Global Self-Worth for any grade level. There is a suggestion, then, that enrichment education may have a particularly beneficial effect for girls. This conclusion is consistent with that offered by Milgram and Milgram (1976).

Table 2 also presents normative scores for the various SPPC subscores. These have been calculated from scores presented in the Manual (Harter, 1985) for grades 5 through 8. Comparisons between the current sample and these norms based on children in regular classrooms reveal some interesting findings. Pupils from the current sample present somewhat lower scores in Social and Athletic Competence, but substantially higher scores on Scholastic Competence. This result is generally consistent with other reports in the literature (Kelly & Colangelo, 1985; Schneider et al., 1989). The girls in our sample present substantially higher scores on Scholastic Competence and Global Self-Worth than did girls drawn from regular classrooms. This finding, too, is consistent with our suggestion that enrichment type classes may be particularly beneficial for girls.

Correlations Among Subscores

Table 3 presents intercorrelations among the SPPC subscores, including both specific and global scales. Coefficient Alpha values are presented in the diagonal of the table.

Consistent with Byrne and Schneider (1988) and Harter (1985), low-to-moderate correlations were obtained among the five specific subscales. These ranged from a low of -.05 between Physical Appearance and Athletic Competence to a high of .42 between Scholastic and Social Competence. These moderate correlations lend some support to the validity of the subscores by demonstrating that they are, in fact, relatively independent of one another. In addition, the Coefficient Alpha values indicate satisfactory levels of internal consistency for the various scales. (Relations between subscores and the global factor, General Self-Worth, are discussed in the next section.)

We also examined correlations among the five specific subscales across the four grade levels. There was, however, no support for Harter's (1985) finding that the five areas become increasingly differentiated across ages. In fact, average correlations among the five subscales were quite close for each of the grade levels: fifth grade, mean r = .22; sixth grade, mean i- = .20; seventh grade, mean i- = .1 7; eighth grade, mean i- = .22. One possible explanation for these discrepant findings is that intellectually select children may evolve a differentiated self system earlier than do less gifted children; that is, of course, a speculation requiring additional investigation.

Multiple Regression Analyses

Multiple regression analyses were performed in an effort to explore the components of global self-worth. The first set of these analyses employed the five specific subscores of SPPC as predictor variables and the Global Self-Worth factor as the criterion variable. These analyses were performed separately for the entire group of students, by gender and by grade level. Table 4 contains a summary of Beta weights and multiple R values,

For the total group of subjects, the Social Competence factor received the heaviest weight in the equation, followed by Physical Appearance and Scholastic Competence. The heavy weights associated with the social and appearance factors are consistent with the findings of Byrne and Schneider (1988) and Harter (1985). The significant contribution of scholastic competence to the global index is, however, not typical of pupils in regular classrooms and is probably unique to gifted pupils for whom academic excellence has been emphasized and rewarded.

Male and female pupils present similar patterns of relations between specific subscores and Global Self-Worth. It is worth noting, though, that Scholastic Competence receives somewhat more weight for girls than for boys, suggesting that school achievement may be of somewhat more significance for girls at these age levels.

The pattern of social competence and physical appearance receiving the greatest weights in the prediction of Global Self-Worth holds, it can be seen from Table 3, for each grade level except 8. In the latter case physical appearance and scholastic competence received the highest weights.

Table 5 presents Beta weights and multiple R values from a second set of multiple regression analyses. The criterion variable is again Global Self-Worth, but the predictor variables in this case are the four subscores of the SRBCSS instrument: Learning, Motivation, Creativity, and Leadership. This analysis provides us with information on the way in which teacher-perceived strengths of pupils relate to self-esteem in the pupil.

The multiple regression analysis based on the total sample of students revealed a significant weight associated with the Leadership score and the Creativity score approaching significance, p < .05. Higher ratings on leadership were associated with higher levels of self-worth. This result is consistent with the findings from the previous analysis to the extent that this leadership factor parallels the Social Competence factor from the SPPC. The negative relation between the Creativity score and Global Self-Worth, while only approaching significance, is of some interest. This Creativity subscale includes items reflecting somewhat positive attributes: high levels of curiosity, sensitivity, and divergent thinking abilities. It also contains attributes, however, that may not be particularly valued at this age level: radical and spirited in disagreement," "nonconforming," "high risk taker," "unwilling to accept authoritarian pronouncements without critical examination."

Psychological perspectives on the self, . Vol. 3 (pp. 139-18 1). Hillsdale, NJ: Lawrence Erlbaum.

Kelly, K., & Colangelo, N. (1985). Academic and social self-concepts of gifted, general, and special students. Exceptional Children, 50, 551-554.

Milgram, R. M., & Milgram, N. A. (1976). Personality characteristics of gifted Israeli children. Journal of Genetic Psychology, 129, 185-194.

Renzulli, J. S., Hartman, R. K., & Callahan, C. M. (1971). Teacher identification of superior students. Exceptional Children, 38, 211-214.

Schneider, B. H. 1987). The gifted child in peer group perspective. New York: Springer-Verlag.

Schneider, B. H., Clegg, M. R., Byrne, B. M., Ledingham, J. E., & Crombie, G. (1989). Social relations of gifted children as a function of age and school program. Journal of Educational Psychology, 81, 48-55.

Shavelson, R. J., Hubner, J. J., & Stanton, G. C. (1976). Self-concept: Validation of construct interpretations. Review of Educational Research, 46, 407-441.

Winne, P. H., & Marx, R. W. (1981, April). Convergent and discriminant validity in self-concept measurement. Paper presented at the annual meeting of the American Educational Research Association, Los Angeles.


ROBERT D. HOGE is a Professor and ROBERT MCSHEFFREY is a Graduate Student in the Department of Psychology at Carleton University, Ottawa, Ontario, Canada.

We thank the teachers and principals who participated in this study. We also extend our appreciation to Hazel Bowen, Enrichment Consultant, Ottawa Board of Education; and Kelly Sims, Cheryl McDougall, and Jim Vyse. Requests for reprints should be sent to Robert D. Hoge, Department of Psychology, Carleton University, Ottawa, Ontario, K I S 5B6, Canada.

Manuscript received March 1989; revision accepted July 1989.

Exceptional Children, Vol. 57, No. 3, pp. 238-245. [c] 1990 The Council for Exceptional Children.
COPYRIGHT 1990 Council for Exceptional Children
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Hoge, Robert D.; McSheffrey, Robert
Publication:Exceptional Children
Date:Dec 1, 1990
Previous Article:Apprenticeship and intensive training of consulting teachers: a naturalistic study.
Next Article:Curriculum-based measurement and two models of follow-up consultation.

Related Articles
Relationship between personality adjustment and high intelligence: Terman versus Hollingworth.
Feelings and attitudes of gifted students.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters