The Impact of Locus of Control, Self-Esteem, and Self-Efficacy on the Academic Performance of University Students: Evidence from Public Sector Universities in the Southern Districts of Khyber Pakhtunkhwa, Pakistan

Abstract

This study aims at exploring the effect of the combined and relative influence of three basic psychosocial constructs (locus of control, self-esteem and self-efficacy) on academic performance of undergraduate university students studying in the five public sector universities of the southern districts of Khyber Pakhtunkhwa, Pakistan (Karak, Lakki Marwat, Kohat, Bannu and Dera Ismail Khan). Each of these constructs has been individually studied for its relationship with student achievement in international literature, but few empirical studies have examined the joint predictive power of these constructs in a single structural model in a resource-poor, South Asian context in higher education. The study describes and examines a structural model that is based on Rotter's Social Learning Theory Locus of Control, Bandura's Social Cognitive Theory Self-efficacy, and Socio-meter Theory Self-esteem, in which Locus of Control, Self-esteem, and Self-efficacy are proposed to be direct precursors of academic performance. The research design was cross-sectional, quantitative, positivist and the data was collected from a target population of 3122 registered students from five universities and a working sample of 343 registered students used, whose data was usable. Standardized instruments that were used included Rotter's Internal-External Locus of Control Scale, the Rosenberg Self-Esteem Scale, and Schwarzer and Jerusalem's General Self-Efficacy Scale as well as a self-reported cumulative grade point average (CGPA) measure of academic performance. In line with contemporary methodological literature, data were analysed using a two-stage approach of measurement and structural model assessment, which was achieved using Partial Least Squares Structural Equation Modelling (PLS-SEM) in SmartPLS 4. The indicators of the measurement model showed acceptable reliability, internal consistency, convergent and discriminant validity. Results from the structural model are expected to show that all three of the variables (internal locus of control, self-esteem, and self-efficacy) have a statistically significant positive impact on academic performance, with the self-efficacy variable expected to be the strongest of the three in line with the evidence from the wider literature in the form of a meta-analysis. The study has implications for the literature on psychology of education by simultaneously modeling three well-established and seldom tested predictors in an underrepresented context, as well as in the context of student counselling services, faculty advising, and curriculum design in Pakistani higher education institutions. The cross-sectional design and self-report measurement limitations and directions for further, longitudinal and experimental research are discussed.