Concurrent and Longitudinal Predictors of Self-Esteem in Elementary and Middle School Girls

BACKGROUND: The purpose of this study was to analyze the effects of factors related to self-esteem, both cross-sectionally and longitudinally, among 2 cohorts of girls over a period of 4 years, from elementary through middle school. METHODS: A multiethnic sample of 656 elementary school girls recruited from 13 schools in Hayward, CA, and Tucson, AZ, was evaluated annually over a 4-year period. The McKnight Risk Factor Survey IV was administered, which consists of 103 questions that assess self-esteem, appearance appraisal, effect of body changes, depressed mood, teasing, school performance, and other factors. In addition, participants' heights and weights were measured. RESULTS: The most important predictor of self-esteem in each grade was appearance appraisal, and, in all grades but the eighth, weight-related teasing by either boys or girls was also a predictor of self-esteem. Teasing about weight was more important than body weight itself in predicting self-esteem. Self-esteem was lower in girls who were teased about their weight, even if they believed that it had no effect on how they felt about themselves. School performance predicted self-esteem in grades 6 and 8. In longitudinal analyses, the participant's current perception of her appearance was more important than her previous level of self-esteem in predicting current self-esteem. CONCLUSIONS: The results of this study support the need to allocate funding to address the issue of teasing as 1 of the stressors in the school environment, and to explore further the important relationship between school performance and self-esteem.