Using Structural Equation Modelling to Understand Predictors of Undergraduate Students’ Academic Performance

This chapter argues that there are many, just many many variables which contribute to academic performance as measured in degree outcome, and, as such, simple bivariate analysis is inappropriate. We use structural equation modelling, and explore the contribution of academic behavioural confidence, to make the point that it does contribute to academic performance, but to a lesser extent than self-efficacy theory argues. We suggest that this is because degree outcome is made up of many efficacy variables, which we argue are better captured overall in academic behavioural confidence.