Hassan’s PhD research aimed to develop comprehensive analytical frameworks for the before-after evaluation of engineering treatments utilising heterogeneous econometric models. In particular, his research endeavoured to achieve four objectives. First, to investigate the effect of different segmentation approaches on the safety effectiveness of engineering treatments. Incorporating the learning from the first objective, his second objective was to propose a framework that accommodates unobserved heterogeneity in the Empirical Bayes approach. Third, he examined the performance of various engineering treatment evaluation techniques using a hypothetical treatment setup. Fourth, he proposed a novel mixed distribution multivariate model that simultaneously estimates multiple crash types while accounting for the correlation in multiple crash types by severity, the unobserved heterogeneity, and preponderance of zeros in evaluating the safety of the whole highway system.