Examining Undergraduate University Programs’ Appropriateness at Employability Outcomes in Jordan and Egypt Using Ordered Forests Method

This study examines university graduates’ ratings of the Bachelor of Science (BSc) program’s appropriateness regarding employability outcomes in the labor market. The study uses crosssectional data from two countries in the Middle East and North Africa (MENA) region, Egypt and Jordan. The graduates’ rating is modelled using the ordered response method, where we apply the ordered forests machine learning technique, and compare its results with the conventional ordered probit model. The results show that differences in the education system and labor market opportunity structure influence graduates’ ratings of the appropriateness of the BSc program. Graduates’ skill matching in their jobs is found to be the strongest factor that impacts the ratings and the satisfaction about the university degree. Policies are required to enhance graduates’ employability. The marginal effects of the ordered probit model can be implied as significant, even when the size of the effect is trivial and has no real impact on the conditional probabilities, but can easily be linked to the relative importance ranks of the covariates in the model, in contrast to the ordered forests.