Development of Domain-Specific Self-Evaluations: A Meta-Analysis of Longitudinal Studies

• In Event: Heterogeneity in Self-Esteem Development: New Theoretical Insights and Longitudinal Evidence

Abstract

When individuals think about whether they are satisfied with their competences and personal characteristics, they may come to very different conclusions. Notwithstanding these individual differences, it is possible that the average positivity of self-evaluations follows a systematic, normative trajectory across childhood, adolescence, and adulthood (Harter, 2006). Moreover, the normative trajectory might be domain-specific; that is, the trajectory might depend on whether self-evaluations refer to the academic domain, social relationships, sports, or physical appearance. Many studies have examined trajectories of domain-specific self-evaluations (also referred to as self-concept or domain-specific self-esteem), both decades ago (e.g., Cole et al., 2001; Eccles et al., 1989; Marsh, 1989) and more recently (e.g., Esnaola et al., 2020; Harris et al., 2018; Kuzucu et al., 2014). However, despite the large body of evidence that has been accumulated, the findings have been inconsistent and have not yet led to any agreement on the normative development of domain-specific self-evaluations. In this situation, meta-analytic methods are ideally suited to gain more robust insights into developmental patterns.

Therefore, this meta-analysis investigated the normative development of domain-specific self-evaluations by synthesizing the available longitudinal data on mean-level change. Eight domains of self-evaluations were assessed: academic abilities, athletic abilities, physical appearance, morality, romantic relationships, social acceptance, mathematics, and verbal abilities. Analyses were based on data from 143 independent samples which included 112,204 participants. Of the studies, 32% used one of the scales developed by Harter, 18% one of the scales by Marsh, and 50% another measure. As the effect size measure, we used the standardized mean change d per year. There was no evidence of publication bias. The mean age associated with effect sizes ranged from 5 to 28 years. To illustrate the patterns of change, we aggregated the point estimates of mean-level change across the observed age range for each of the domains.

The results showed that developmental trajectories of self-evaluations were positive in the domains of academic abilities, social acceptance, and romantic relationships. In contrast, self-evaluations showed negative developmental trajectories in the domains of morality, mathematics, and verbal abilities. Little mean-level change was observed for self-evaluations of physical appearance and athletic abilities. Moderator analyses were conducted for the full set of samples and for the subset of samples between ages 10 and 16 years. In the moderator analyses, the pattern of findings held across demographic characteristics of the samples, including gender and birth cohort. Thus, the results suggest that girls/women and boys/men do not differ in the typical patterns of change in self-evaluations. Similarly, the results indicate that the normative trajectories of self-evaluations have not changed over the generations born since the mid-20th century. The meta-analytic findings suggest that the notion that self-evaluations generally show a substantial decline in the transition from early to middle childhood should be revised. Also, the findings did not support the notion that self-evaluations reach a critical low point in many domains in early adolescence.

Authors

• Ulrich Orth, Universität Bern
  Presenting Author

• Laura C. Dapp, University of Bern
  Non-Presenting Author

• Ruth Yasemin Erol, University of Bern
  Non-Presenting Author

• Samantha Krauss, Universität Bern
  Non-Presenting Author

• Eva C. Luciano, University of Bern
  Non-Presenting Author