The Determinants of University Students Success: a Bivariate Latent Variable Model

The analysis of performance indicators of university students has become of wide interest expecially in Italy where, over the last few decades, graduation rates have been well below the average of both European and OECD countries. This paper proposes an alternative method to jointly estimate the determinants of students academic success, in terms of both potential credits and retention, one year after they rst enrolled and a further analysis to evaluate whether there are any factors signicantly determining the probability of dropping out, once we consider the students potential academic performance ceteris paribus. We implement the algorithm to estimate the parameters of a bivariate latent variable system and then of a conditional mean equation.