Mental health challenges pose significant barriers to academic success and well-being among university students, particularly in engineering and computer science. While prior research has examined factors influencing students' mental well-being, a gap remains in how these factors rank in relative importance and vary across demographic groups. This study identifies and ranks factors predicting mental health changes among engineering and computer science undergraduates in the U.S., and examines how these relationships vary by gender, race/ethnicity, and first-generation status. Using random forest analysis on the nationally representative BPS 12/17 Survey (2011-2014), we found that physical health was consistently the most important predictor of mental health outcomes, followed by academic preparation, institutional characteristics, social support, and psychological factors that varied in importance. Sub-population analyses revealed significant heterogeneity and distinct additional challenges faced by marginalised and first-generation students navigating compounding intersectional pressures.
Xu et al. (Fri,) studied this question.