Structure and Change in Self-Concept During Adolescence

Canadian Journal of Behavioural Science, Apr 2005 by Shapka, Jennifer D, Keating, Daniel P

Abstract

We investigated the longitudinal changes in multiple domains of self-concept over a two-year period, for 518 high school students, using a multioccasion-multicohort design. We were interested in identifying gender and age effects, as well as in examining the hierarchical nature of self-concept by determining the domains of self-concept that are most predictive of one's overall sense of self-worth. Using Harter's Self-Perception Profile, we found that most domains of self-concept increase with age, although perceived scholastic competence decreases. There were several gender differences, all of which replicated known discrepancies in the literature. In addition, perceptions of appearance were most closely tied to general self-worth, and this pattern was stable over time. As well, we found that current and changing levels of the various domains of self-concept were the best predictors of change that occurred in overall self-worth. Finally, we found that weighting the various domains of self-concept by the importance that each individual places on that domain did not improve the predictive power of specific domains to general self-worth.

R�sum�

Nous avons examin� les changements longitudinaux dans plusieurs domaines du concept de soi pendant une p�riode de deux ans, aupr�s de 518 �tudiants de niveau secondaire � l'aide d'une conception multicohorte � occasions multiples. Nous cherchions � identifier les effets du sexe et de l'�ge, ainsi qu'� examiner la nature hi�rarchique du concept de soi, en �tablissant les domaines du concept de soi qui sont les plus susceptibles de pr�dire le sentiment global de confiance en soi qu'�prouv� une personne. Au moyen du profil de perception de soi de Harter, nous avons d�couvert que la plupart des domaines du concept de soi se renforcent avec l'�ge, m�me si les comp�tences scolaires per�ues d�croissent. Nous avons observ� plusieurs diff�rences entre les filles et les gar�ons, qui reproduisaient tous les �carts connus dans les ouvrages scientifiques sur le sujet. En outre, les perceptions relatives � l'apparence �taient les plus �troitement reli�es � la confiance en soi et ce mod�le �tait stable au fil du temps. De plus, nous avons constat� que les niveaux actuels et changeants des divers domaines du concept de soi �taient les meilleurs pr�dicteurs du changement qui se produit dans la confiance en soi globale. Pour finir, nous avons d�couvert que, en pond�rant les divers domaines du concept de soi avec l'importance qu'attribu� chaque personne � ce dernier, nous ne voyions pas une am�lioration de l'efficacit� pr�dictive de domaines particuliers sur la confiance en soi globale.

Self-evaluations influence the kinds of activities one engages in, the effort and motivation expended, and the likelihood of future perseverance in that activity. Much attention has been focused on elucidating exactly how self-perceptions influence different kinds of outcomes - including both direct and indirect links (e.g., Cairns, McWhirter, Duffy, & Barry, 1990; Harter, 1999; Marsh, 1989; Marsh & Hattie, 1996). There continue to be gaps, however, in our understanding of gender differences, age-related changes, and the relationship between different domains of self-evaluation (Grain, 1996). Laying out this basic groundwork is a necessary step before other substantive questions can be adequately addressed. This goal is particularly paramount for adolescent development. Given the hypothesized links between low self-concept and maladaptive outcomes (such as low academic achievement and depression), in conjunction with the heightened salience of self-perceptions for adolescents, this age group is often viewed as the most susceptible to the negative ramifications associated with low self-esteem (Harter; Marsh). This study thus utilizes Harter's Self-Perception Profile to provide an empirical understanding of how self-concept changes over the course of high school, as a function of gender and age. In addition, we attempt to identify the stability and change in the hierarchical structure of self-concept, and to understand the developmental mechanisms underlying self-concept instability.

Until recently, dominant practice has been to measure self-concept using a single score, derived either by summing or averaging positive or negative perceptions of worth across several facets of life (Marsh & Hattie, 1996). However, Wylie's work (Wylie, 1979) clearly showed that aggregations across domains of self-concept are not appropriate because they do not incorporate individual differences in the valuing and weighting of each domain. For example, low perceptions of self-worth in a particular domain may not be debilitating to overall self-worth if the person has discounted the importance of that realm of life (Harter, 1999). Recent research has emphasized the limitations of global self-esteem scores, which do not provide an accurate representation of a person's general sense of himself (Cairns et al., 1990; Harter). Current evidence suggests that self-perceptions are much more complex and consist of varying perceptions across several dimensions. More recent research, therefore, examines self-concept from a multifaceted, domain-specific perspective, with several domains of evaluation, ranging from physical appearance and athletic competence to academic perceptions and job competence (Harter; Marsh, 1989).