Structural equation modelling of academic performance in statistics degree programme: A case study of University of Ibadan

Abstract

This paper is concerned with the structural equation modelling of the academic performance in Statistics degree programme.Structural equation modelling is a class of methodologies that represent hypotheses about variances and covariances ofobserved data in terms of structural parameters defined by a theoretical model. It takes a confirmatory approach to themultivariate analysis of structural theory which specified the causal relations among the multiple observed variables. Thecausal pattern of intervariable relations within the theory was specified a priori. Structural equation modelling evaluatedtwo models – the measurement model and the structural model. The measurement model related observed responses orindicators to latent variables (confirmatory factor analysis), while the structural model specified relations among latentvariables (path analysis). The direct and indirect effects of personal factor, psychological factor, institutional/environmentalfactor, family characteristics, social and religious factor on the academic performance of students were studied. Theacademic performance of students remains a top priority for students, parents, educators, researchers, administrators(management) and government. The extent to which the theoretical model is supported by sample data collected fromundergraduate students of the Department of Statistics, University of Ibadan, was determined. There were 37 observedvariables and 7 latent variables. The result indicated that the attitude of students towards learning had positive directimpact on academic performance, while psychological factor had negative influence on academic performance. Also, thestructural equation model analysis with four basic fit indices suggested a reasonable model-data fit.