(PS7-67) Psychometric Evaluation of a Questionnaire Measuring the Degree to Which the College Environment Fosters College Students’ Strengths: Implications for Belongingness & Depression

Abstract Text:

Introduction: Approximately 18.5% of new college students meet criteria for past-year major depressive disorders (Auerbach et al., 2018), and the transition period is particularly daunting for mental health (Mehr & Daltry, 2016). Using one’s character strengths appears to buffer against pessimistic attribution styles and student depression (Lee et al., 2020). In this study, we evaluated the psychometric properties of the Fostering Strengths in College Survey (FSCS): a new 4-item questionnaire measuring the extent to which students perceive that the college environment fosters use of their natural strengths and abilities. We hypothesized that the 4 FSCF items could be captured by a unidimensional latent factor that would be stable over time (longitudinal invariance) and predict belongingness and depressive symptoms (criterion validity).

Method: Data were drawn from a prospective cohort study (N=276) evaluating mental health outcomes among incoming students (first-time students and transfers) during their first semester at a public university. Participants completed 6 assessments: 1 prior to the start of the semester and 5 during the fall semester (Sep – Dec). The FSCS was completed monthly at timepoints 4-6 (Oct, Nov, and Dec). Depression (PROMIS Depression short form; Irwin et al., 2012) and belongingness (Sense of Belonging Scale; Yorke, 2016) measures were completed at all time points. We used longitudinal structural equation modeling to evaluate the tenability and stability of the one-factor FSCS model. Models were estimated using maximum likelihood with robust standard errors and scaled chi-square (x2) tests. Configural, metric (weak), scalar (strong), and uniqueness (strict) invariance models were compared using likelihood ratio tests.

Results: All models showed minimal evidence of model-data discrepancy (non-significant x2 tests). One FSCS item did not significantly load onto the latent factor and was eliminated. The 3-item longitudinal measurement model fit the data well (x2[31]= 32.56, p=.39) and the latent factor had excellent reliability (McDonald’s omega=0.91, 95% CI [0.89, 0.92]). There was evidence of violation of scalar invariance: a model allowing factor indicator intercepts to vary over time was superior to a model holding intercepts constant. However, the degree of model enhancement achieved by allowing time-varying intercepts was negligible: 0.007 points improvement on the Tucker-Lewis Index (TLI). Further, the strict invariance model (loadings, intercepts, & errors held constant) had the best TLI and Akaike Information Criterion scores. Cross-lagged panel models showed strong contemporaneous correlations between FSCS scores and belongingness and depressive symptoms in expected directions. Further, T4 FSCS scores predicted T5 criterion outcomes (belongingness and depressive symptoms) controlling for prior levels of the criterion outcomes.

Conclusion: A 3-item version of the FSCS shows initial promise in studying the mental health consequences of student perceptions that the college environment fosters their character strengths during the transition to college. New data will be collected in the coming year allowing for further validation and scale refinement.