2026 ASEE Annual Conference & Exposition

In this full empirical study, we reveal latent mental health profiles among engineering students using Gaussian Mixture Modeling (GMM). Similarly, we validate clusters through multiple statistical indices and modeling approaches. Previous studies show the increasing rates of mental health distress among engineering students, which affects academic performance, student retention, and overall well-being. Clustering methods such as GMM allow us to segment students into subgroups based on their mental health profiles. Findings may help us to develop targeted interventions per subgroup to support and improve engineering students' well-being. To conduct this work, we examined 37,673 anonymous engineering student records from the Healthy Minds Study (years 2020-2024). Our key measures were the Flourishing Scale, GAD-7 anxiety, and PHQ-9 depression instruments.

We gathered validity evidence for these instruments with the engineering population using exploratory and confirmatory factor analysis before proceeding with clustering. For clustering, we first determined the optimal number of clusters using the elbow and silhouette method and inspection of the within-cluster sum of squares across k=1–10, which was k=3. Then we fit GMM via the mclust package in RStudio. To ensure the validity of our cluster solution, we ran each model 1,000 times for each covariance structure (VVV, EEE, VEV, VII, and EII). Results revealed three distinct profiles: 1) a cluster with high flourishing and low anxiety and depression, 2) a second cluster with high flourishing and moderate anxiety and depression, and 3) a third cluster with low flourishing and high anxiety and depression.

Our clustering process, part of a broader effort to understand what factors are protective of engineering student wellness, helps us begin to understand how engineering students experience mental health. This understanding may help us inform wellness programs and screening procedures and plan for the development of targeted support services. Future work will expand analyses to marginalized populations and incorporate longitudinal and qualitative approaches to better capture the complexity of engineering students’ mental health.