Mental Health in Graduate & Professional Education: Preliminary Findings from a National, Multidisciplinary Study

• In Event: Excellence in Education Research: Early-Career Scholars and Their Work

Abstract

Overview:
This poster will present preliminary findings from the first national study of graduate student depression and anxiety, focusing on the role of academic disciplines as intellectual contexts and learning environments. Data for the poster are drawn from the Healthy Minds Study, which utilizes validated screening measures for depression and anxiety and which samples 20,374 graduate students from 111 universities. This rich sample will offer the first analysis of individual risk factors for mental illness in the graduate student population, and variation in prevalence rates for anxiety and depression across fields of study. In the poster, I report descriptive statistics, including a hierarchical cluster analysis using innovative data visualization methods, as well as multivariate logistic regressions and associated analyses of marginal effects for populations of interest.

Prior Literature:
Little previous research has been undertaken on the mental health of graduate students. The empirical studies that have been conducted typically focus on single disciplines or professional programs (e.g., mental health of students in MD or JD programs) or multiple disciplines within a single university. This work has identified characteristics of the graduate student experience-- such as isolating work, inconsistent quality of mentoring, uneven support from peers and family, impostorism, and financial stress—that place students at risk for anxiety and depression.

It is as yet unclear, however, whether sub-populations of graduate students (either by personal background or field of study) differentially bear these risks and/or whether racial and gender disparities in anxiety and depression may vary across disciplines in the graduate student population. Such information would be vital for faculty and administrators, given that depression and anxiety are both negatively associated with persistence.

Theory:
Role strain theory relates qualities of the learning environment, student characteristics, and mental health outcomes. Bowman’s (2013) posits, ““Chronic environmental risk factors increase personal vulnerabilities, which, in turn, reduce psychological wellbeing” (p. 301). For example, I analyze whether anxiety and depression are associated with graduate student perceptions of competitiveness in their courses, experiences with discrimination and support, and underrepresented/marginalized identity status.

To understand the possibility of and reasons for heterogeneity in mental health disparities and/or risk factors for depression and anxiety across fields, I draw from theories of disciplinary cultures. This is an area of prior research for me, and scholarship to date has elucidated specific ways in which disciplinary knowledge, community norms, and approaches to mentoring may vary across fields. For example, strong paradigm disciplines have more centralized, hierarchical knowledge structures that manifest in the expectations placed upon students and scholars, as well as the nature of learning and apprenticeship.

Methods:
Although my broader NAEd/Spencer postdoctoral fellowship study is a sequential, mixed methods project, the poster will present only data from the quantitative side of the project.

Data come from the 2007-2012 administrations of the Healthy Minds Study (HMS), an annual web-based survey examining mental health and related issues (e.g., depression, anxiety, and substance use), as well as service utilization among postsecondary students. Depression and anxiety are measured using standard screening instruments that have been validated with diverse populations (Spitzer, Kroeke, & Williams, 1999). Other measures are derived from HMS survey items, and have undergone sensitivity analyses to ensure reliable operationalization. The analytic sample consists only of graduate and professional students.

Analyses:
Descriptive analyses examined the prevalence of depression and anxiety by student background and field of study. Through hierarchical cluster analysis, I related and jointly displayed the depression and anxiety prevalence rates by field. The poster promises to be an especially constructive medium for sharing these results.

The dependent variables are both binary; therefore, I conducted separate, multivariate logistic regression analyses to estimate the probabilities that characteristics associated learning environments predict anxiety and depression, controlling for student-level covariates. To assess support for role strain theory would require causal inference, but in the absence of an opportunity for an experimental design, I use regression adjustment techniques to reduce bias in beta estimates. These adjustments included a dichotomous variable for each university as a campus-level fixed-effects estimator and a thorough set of covariates to minimize violations of the strict exogeneity assumption. Estimates, which are reported as marginal effects, represent the conditional correlations between each of the key independent variables and the two mental health outcomes. To examine heterogeneous relationships across student background and field of study, I test for interaction effects and report the results as marginal effects.

Key multivariate findings:
• Financial struggle is the top predictor of both anxiety and depression, across fields, controlling for 40 covariates. Compared to those who say finances are “tight” or “not a problem,” graduate students who report that finances are a “struggle” have 2.3 times higher odds of screening positive for depression and 3.3 times higher odds of screening positive for anxiety. Support from family and friends reduces risk of both anxiety and depression, for students who report financial struggle.

• Racial discrimination Across fields, frequent racial discrimination is the second strongest correlate of anxiety and depression, all else equal. Compared to their peers, graduate students who report experiencing frequent racial discrimination have a 3.01 times higher odds of screening positive for anxiety and a 1.95 times higher odds of screening positive for depression.

• LGBTQ graduate students are at greater risk for both anxiety and depression than any other sub-population surveyed.

• Interactions for female, Black, and Latino x STEM yielded non-significant relationships with both anxiety and depression.

Author

• Julie Renee Posselt, University of Michigan