An Examination of self-concept in academically gifted adolescents: do gender differences occur?

"Too many gifted girls, ebullient, confident, and filled with high aspirations, simply and quietly disappear" (Klein & Zehms, 1996, p. 30). One perspective is that gifted females tend to disappear as they progress through the education system-and ultimately as they enter adulthood due to gender-biased, eminence-based definitions of giftedness. However, Silverman (1993) maintains that the progressive loss of talent in females is a result of socialization practices that undermine gifted females' self-confidence and aspirations. Research into the relationship between self-concept and achievement patterns in females has shown that academic self-concept influences academic success, career choice, and test performance (Olshen & Matthews, 1987).

Within New Zealand culture, little investigation of these issues, particularly as they relate to gender differences among the gifted, has been undertaken. While studies in the U.S. have demonstrated a decline in self-concept as females progress through schooling (Klein & Zehms, 1996), pronounced achievement differences between males and females (Heller & Zeigler, 1996, p. 201), and an inability to achieve adolescent aspirations upon reaching adulthood (Hollinger & Fleming, 1992), do these generalizations cross cultural boundaries? This study investigated whether gender differences in self-concept occur in a small sample of academically gifted adolescents in one New Zealand provincial co-educational secondary school.

The Dilemma

Adolescence is a critical time in the development of self-concept. The definition used for self-confidence in this article; a multidimensional model of self-concept portrayed as a set of related but independent cognitive, perceptual, affective, and evaluative dimensions; served as the theoretical base for this study (Harter, 1986, cited in Hoge & Renzulli, 1991). Aspects of self-concept that were investigated included external dimensions of physical, moral, personal, family, social, academic/work, and the internal aspects of identity satisfaction and behavior as measured by the Tennessee Self-Concept Scale: 2. By examining the kaleidoscope of self-concept as individual pieces, rather than a total image, the intent of this study was to better explain gender differences.

A multidimensional perspective of self-concept recognizes that development occurs from a combination of the internal factor of competence/importance discrepancy and the external factor of social support/positive regard (Harter, 1986, cited in Hoge & Renzulli, 1991). There are gender differences in how important it is to be competent in a particular area, with females rating social areas as more important than males (Kerr, Colangelo, & Gaeth, 1988). Additionally, gifted females receive many mixed messages from parents, teachers, peers, and society (Reis, 1987) -- possibly more so than their male counterparts.

These mixed messages might well leave students in a quandary. Harters' (1986, cited in Hoge & Renzulli, 1991) social support/positive regard process refers to how individuals use the reactions of others to assess performance. Gifted individuals' heightened sensitivity combined with perfectionism means they imagine others as being very critical of them. These beliefs may lead to gifted children developing a negative self-concept (Silverman, 1993).

This may well be true for both sexes, as all adolescents receive pressure from peers to conform to sex role stereotypes (Dusek & Flaherty, 1981). However, adolescent females use social interaction to determine the quality and acceptability of their achievements. Therefore, a bright female may see her giftedness as making her unpopular with her peers. This fear of success may lead to females underachieving so they do not receive the social penalties of appearing to be too bright. Underachievement also means they have more time to pursue social activities, which they rate more highly than academic achievement. In addition, females perceive themselves as getting less support from their classmates and friends than males. These factors may lead to gifted females feeling less competent and having a lower self-concept than gifted males (Kramer, 1985, cited in VanTassel-Baska, Olszewski-Kubilius, & Kulieke, 1994).

Dweck (1986) found that females had a tendency towards lower expectations (unrelated to actual competence), were debilitated under failure and attributed their failure to lack of ability. This can result in females developing a fear of failure and significantly higher levels of anxiety than gifted males (Tong & Yewchuk, 1996).

Loeb and Jay (1987) suggest that secondary schools are more supportive of the masculine achievement style and this, combined with fear of success and fear of failure, may lead to a lower self-concept in adolescent females. This achievement style is in contrast to primary schooling where achievement occurs through conformance, a more congruent match to the female ideal of being well behaved, conscientious, and obedient. This feminized schooling, however, does not suit the traditional male role of being aggressive, self-reliant, and individualistic. Therefore, academic success can lead to self-doubt in preadolescent boys. Loeb and Jays' (1987) work on the impact of age and schooling style may help explain why varying results have occurred in studies of gender differences in the self-concept of gifted students. This variation in results is one reason that it was important to conduct this study, to determine in what ways the self-concept of academically gifted male and female students at a large provincial secondary school in New Zealand differed.

Methodology

A quasi-experimental research design was utilized, with participants assigned on the basis of gender rather than randomly. The sample of forty 15-, 16-, 17-, and 18-year-old academically able students was selected from a large rural co-educational high school of 1,000 students. Equal numbers of males and females at each level were selected on the basis of entrance testing, examination results, and teacher referral. The aim was to identify the highest academic 5% of the school population.

The independent variable in the study was gender. The dependent variable was self- concept as measured on the Tennessee Self-Concept Scale: Second Edition (TSCS: 2) (Fitts & Warren, 1996). The advantage of using a standardized test like the TSCS: 2 was that its reliability and validity had been extensively evaluated and recognized. In order to control for intervening variables, all participants sat for the test at the same time and in the same location. Instructions contained in the testing manual were used to ensure the procedure was standardized. Tests were scored for total self-concept. In addition, physical, moral, personal, family, social, academic/work, identity, satisfaction, and behavior scores were calculated to discover whether gender differences occurred on the various test dimensions.

Mean total self-concept scores for the male and female groups were calculated. Male and female means were calculated for each of the sub-scores to see if there were gender differences in particular areas of self-concept. An independent t test was applied to these results to determine if any differences were statistically significant.

Results

Table 1 shows that gender differences did in fact occur. A trend for males to score higher than females on most of the dimensions measured by the TSCS: 2, other than Self-Criticism, Moral Self-Concept and Inconsistent Responding, can be seen. In all other areas, including Identity and Satisfaction with themselves and their Behavior, males scored higher than females.

The mean male score for Total Self-Concept (52.10) was higher than the mean female score (49.50). However, although some of these mean differences between male and female scores, were quite large, only the Academic/Work Self-Concept was statistically significant. The mean female Academic/Work Self-Concept was 50.20 while the mean male score was higher at 53.90.

Discussion

This study showed a trend for males to have higher self-concept in most of the areas measured by the TSCS: 2. The statistically significant lower female score for academic/work may be explained by Harters' social support/positive regard concept (1986, cited in Hoge & Renzulli, 1991). Internal factors such as perfectionism and sensitivity may lead to gifted females viewing others as being critical and less supportive of them (Silverman, 1993; VanTassel-Baska, Olszewski-Kubilius, & Kulieke, 1994). This combined with external pressure to conform to sex role stereotypes could lead to gifted females feeling less competent and having a lower self-concept than gifted males (Schaffer, 1989, cited in Tong & Yewchuk, 1996).

Females feeling less competent about their academic abilities may also influence their subject and career choices beyond school. Fear of failure that contributes to gifted females choosing softer options at school may then limit their opportunities for career advancement and professional success (Davis & Rimm, 1998).

Females scored higher than males on Moral Self-Concept, which measures whether the subject does the right thing. The conformance and compliance required to score highly in this area is consistent with Silverman's (1993) work that suggests that gifted females are good at fitting in and adapting to social situations. Females value social situations and it is important to them they are competent in this area (American Association of University Women [AAUW] Education Foundation, 1992). This leads to their Moral Self-Concept being highly developed. Males do not place such importance on social situations; therefore, their Moral Self-Concept is generally lower than females (Harter, 1986 cited in Hoge & Renzulli, 1991).

Females also scored higher than males on Inconsistent Responding. This indicates a flexibility in thinking and the use of situational ethics by the females. It is also possible that Inconsistent Responding mirrors the ambivalence gifted females have about themselves. They may struggle between maintaining their personal integrity about who they are and what talents they possess, and meeting the expectations others have of them (AAUW, 1992). They have been socialized to fit into sex role stereotypes and, because gifted females are very adaptable, they may underachieve to fit in to social situations (Silverman, 1993).

Delimitations and Limitations

The small size and single site of the sample for this research limit the extent to which the findings could be generalized to other contexts. The nature of the sample means it is possible that students identified in this study may not be considered gifted relative to other similar studies. Hoge and Renzulli (1991) express concerns about the reliability and validity of measuring devices such as the TSCS: 2. However, the reliability and validity procedures used in the TSCS: 2 follow standard psychometric techniques and the test is well regarded internationally (Fitts & Warren, 1996). There were also limitations with respect to the statistical analysis of results. The small number of subjects meant that large differences were required in order for statistical significance to occur.

Conclusion

The disappearance of gifted females is a complex phenomenon that warrants further investigation. This study suggests that gender differences in self-concept do occur in some New Zealand gifted adolescents. The higher levels of self-criticism, flexible situational responding and compliance to sex roles that were identified in this study may all contribute towards females having lower self-concept, and academic self-concept in particular. Lower self-concept may in turn be a factor in female underachievement. Combined with the importance females place on fitting into social situations, this lower self-concept may partially explain why gifted females disappear. Perhaps when we have discovered what is actually happening through studies such as this, we can put measures in place to address the situation.

Table 1 Mean TSCS: 2 Scores
                                  Mean     Mean     t-Test
                                  Female   Male     Obtained
                                  Score    Score    Value

Validity Scores
 Inconsistent Responding  (INC)   46.50   44.80    0.608883
 Self-Criticism            (SC)   49.45   45.80    1.347493
 Faking Good               (FG)   46.85   49.90    1.26865
 Response Distribution     (RD)   41.45   43.60    0.898929

Summary Scores
 Total Self-Concept       (TOT)   49.50   52.10    1.28044
 Conflict                 (CON)   44.45   48.00    1.38935

Self-Concept Scales
 Physical                 (PHY)   44.80   46.60    0.68867
 Moral                    (MOR)   53.45   52.60    0.30749
 Personal                 (PER)   46.50   50.15    1.56555
 Family                   (FAM)   52.25   54.65    0.98932
 Social                   (SOC)   49.25   49.15    0.040347
 Academic/Work            (ACA)   50.20   53.90    1.70812 *

Supplementary Scores
 Identity                 (IDN)   46.70   48.35    0.65052
 Satisfaction             (SAT)   49.20   52.55    1.46976
 Behavior                 (BHV)   49.55   50.80    0.42624

* Statistically Significant at
0.10 level. Critical Value 1,645

Manuscript submitted March, 2000. Revision accepted February, 2001.

REFERENCES

American Association of University Women Education Foundation. (1992). The AAUW report: How schools shortchange girls. U.S.: Wellesley College Center for Research on Women.

Davis, G., & Rimm, S. (1998). Education of the gifted and talented (4th Ed.). Boston: Allyn and Bacon.

Dusek, J., & Flaherty, J. (1981). The development of the self-concept during the adolescent years. Chicago: University of Chicago Press.

Dweck, C. (1986). Motivational processes affecting learning. American Psychologist, 41, 1040-1048.

Fitts, W., & Warren, W. (1996). Tennessee Self-Concept Scale TSCS: 2 Manual (2nd Ed.). USA: WPS.

Heller, K. A., & Zeigler, A. (1996). Gender differences in mathematics and the sciences: Can attributional retraining improve the performance of gifted females? Gifted Child Quarterly, 40, 200-209.

Hoge, R. D., & Renzulli, J. S. (1991). Self-Concept and the gifted child. Research-Based decision making series. Connecticut: The National Research Center on the Gifted and Talented.

Hollinger, C. L., & Fleming, E. S. (1992). A longitudinal examination of life choices of gifted and talented young women. Gifted Child Quarterly, 36, 207-212.

Kerr, B., Colangelo, N., & Gaeth, J. (1988). Gifted adolescents attitudes towards their giftedness. Gifted Child Quarterly, 32, 245-247.

Klein, A. G., & Zehms, D. (1996). Self-concept and gifted girls: A cross sectional study of intellectually gifted females in grades 3, 5, 8. Roeper Review, 19, 30-33.

Loeb, R., & Jay, G. (1987). Self-concept in gifted children: Differential input in girls and boys. Gifted Child Quarterly, 31, 9-14.

Olshen, S., & Matthews, D. (1987). The disappearance of giftedness in girls: An intervention strategy. Roeper Review, 9, 251-254.

Reis, S. M. (1987). We can't change what we don't recognize: Understanding the special needs of gifted females. Gifted Child Quarterly, 31, 83-89.

Silverman, L. (Ed.). (1993). Counseling the gifted and talented. Denver: Love Publishing.

Tong, J., & Yewchuk, C. (1996). Self concept and sex role orientation in gifted high school students. Gifted Child Quarterly, 40, 15-23.

VanTassel-Baska, J., Olszewski-Kubilius, P., & Kulieke, M. (1994). A study of self-concept and social support in advantaged and disadvantaged seventh and eighth grade gifted students. Roeper Review, 16, 186-191.

Ainsley Luscombe is a Guidance Counselor at Hawera High School. She graduated in 1999 with a Master of Counselling (Distinction) from Massey University, New Zealand. Dr. Tracy Riley is a Senior Lecturer in gifted and talented education at Massey University.