Predictors of academic achievement and academic self-concept: a longitudinal perspective

Background. A large body of research supports the existence of an academic achievement - self-concept relationship. Path analyses mostly show academic achievement to be causally predominant in this relationship. Aims. This study aims to replicate these findings among Belgian primary school-age children, controlling for a number of other relevant variables such as school commitment, parental socio-economic status (SES), gender and teacher expectations. The longitudinal design of this study makes it particularly suited for this purpose. Method. Data were analysed using correlation, regressions of time 1 independent variables on time 2 dependent variables, multiple classification analysis and path analysis using LISREL. Results. Academic self-concept and academic achievement were found to be strong predictors of one another, even controlling for other variables and stability of both over time. The only other variable to enter significantly into regressions on time 2 academic achievement and academic self-concept was parental socio-economic status (SES). Both the regression analyses and path analysis undertaken specifically to test causal predominance found achievement to be causally predominant in this sample. Conclusion. The fact that academic achievement was causally predominant over academic self-concept, that global self-esteem was not a significant predictor of achievement, and the significant contribution of parental SES to achievement all suggest that self-esteem enhancement in itself cannot be a solution to the problem of academic failure.