Relations of academic and general self-esteem to school achievement

Abstract

The study demonstrates on a nationally representative sample of Estonian students and university applicants (N = 4572) that although self-reported academic self-esteem is a strong and accurate predictor of school achievement, additionally rather low, not high, general self-esteem is a significant predictor of superior school performance when academic self-esteem and multicollinearity is controlled for. Two compensatory mechanisms—defensive pessimism and self-protective enhancement—may explain the paradox of low self-esteem: academically successful students have a more critical view of themselves and students with more modest academic abilities compensate for their academic under-achievement by elevating their general self-esteem. Children start to use self-protective enhancement but from age 12 to 14 they also start using defensive pessimism to protect themselves from the consequences of failure.

Introduction

Despite popular beliefs that high self-esteem facilitates academic achievement, only a modest correlation was discovered between general self-esteem and school performance (Byrne, 1984, Byrne and Shavelson, 1986, Hansford and Hattie, 1982, Marsh and Yeung, 1998, Mintz and Muller, 1977). A recent meta-analysis estimated that the effect of the favourable influence of positive self-beliefs on academic achievement is modest and the overall estimated relation is about .08 (Valentine, Dubois, & Cooper, 2004). Also longitudinal studies do not point to any significant role for general self-esteem in the advancing of academic performance (Bachman and O’Malley, 1977, Byrne, 1986, Helmke and van Aken, 1995, Maruyama et al., 1981, Pottebaum et al., 1986, Skaalvik and Hagtveg, 1990). Educational practices attempting to boost children’s self-worth have demonstrated only limited and temporal effects unless they targeted specific self-concept domains (O’Mara et al., 2006). Based on these observations, Baumeister and his colleagues (2003) concluded that generalized evaluation of self-worth has no significant impact on the subsequent academic achievement (see Marsh & Craven, 2006, for a different opinion).

One possible explanation to the absence of a strong association between general self-esteem and academic performance is the specificity matching principle: in order to expect a strong or moderate correlation the specificity of predictors and criteria should be matched (Swann, Chang-Schneider, & McClarty, 2007). Because general self-esteem is unspecific and school performance rather narrowly defined, it may seem unlikely that school grades can be predicted from what people think or at least report about their general self-worth.

Not disputing the specificity matching principle, it is relevant to recognize that self-concept is a multidimensional construct (cf. Shavelson, Hubner, & Stanton, 1976) in which general and academic self-esteem are two distinguishable components with two dissimilar functions. While general self-esteem appears to be heavily affective in nature and tends to be associated with overall psychological well-being, specific self-esteem—that is self-evaluations in narrowly defined domains like school performance—appears to have a more cognitive component and tends to be more strongly associated with behaviour or behavioural outcomes (Baumeister et al., 2003, Rosenberg et al., 1995). This means, in particular, that general self-esteem is only partly based on the student’s academic self-esteem defined as evaluation of their school performance while self-assessment of their academic abilities is sufficiently accurate to reflect their achievement at school. Indeed, a meta-analysis has shown that specific academic self-concepts provide much better prediction of academic achievements than global self-esteem (Hansford & Hattie, 1982).

The main goal of this study is to demonstrate a multidimensional character of self-concept by showing reciprocation between general and academic self-esteem in their joint prediction of school achievement.