Poster #91 - Predictors of Mental Health Among Undergraduate Engineering and Computer Science Students

• In Event: Saturday Poster Luncheon

Abstract

Mental health concerns have appeared as critical barriers to academic persistence and success in
higher education, particularly in high-intensity disciplines such as engineering and computer
science (Eng and CS). However, there is a lack of comprehensive investigation about the relative
importance of contributing factors and how their influence differs across demographic
subgroups, and how can higher education institutions develop more systematic support policies
that take into account these factors and address the diverse backgrounds of students. This study
adopts the Demand-Control-Support Model and Minority Stress model and seeks to address
these gaps by identifying and ranking predictors of mental health changes among Eng and CS
undergraduates, with particular attention to gender, race/ethnicity, and first-generation status.

Using a random forest analysis - a machine learning approach, we analyzed a dataset comprising
the nationally representative BPS 12/17 Survey, American Community Survey data, and FBI
Hate Crime statistics. Results indicate that physical health is the most critical determinant of
mental health outcomes, followed by academic preparation, institutional environment, and social
support. Importantly, psychological and environmental factors showed varied significance across
subgroups. For example, first-generation and racially minoritized students encountered distinct
patterns of risk, including reduced access to preparatory resources and elevated exposure to
hostile campus climates.

These findings offer actionable insights for policymakers, institutional leaders, and student
support services. They suggest that fostering student well-being requires a dual focus: investing
in physical infrastructure (e.g., health services, preparatory bridge programs) and strengthening
social-psychological supports (e.g., belonging initiatives, inclusive pedagogy). Moreover, the
study underscores the necessity of intersectional frameworks when designing mental health
interventions in Eng and CS education. It ensures that diverse student populations are not left
behind.

Author

• Li Tan, Arizona State University
  Presenting Author