Mental Health Symptom Profiles Over Time: A Three-step Latent Transition Cognitive Diagnosis Modeling Analysis with Covariates

This data repository contains the companion data and R code files for Liang, de la Torre, Larimer, and Mun (2024). Mental health symptom profiles over time: A three-step latent transition cognitive diagnosis modeling analysis with covariates. In M. Stemmler, W. Wiedermann, & F. Huang (Eds.), Dependent data in social sciences research: Forms, issues, and methods of analysis (2nd ed.). New York: Springer.

Abstract

Cognitive diagnostic modeling (CDM) is an item-level analysis that accounts for attribute co-occurrences when characterizing attributes and classifying individuals’ attribute profiles. Tan et al. (2023) provided an application for mental health symptom profiles. The current study extends Tan et al. (2023) to demonstrate how intervention and gender affect transition probabilities from one state to another in a three-step latent transition CDM. The sample used in this study consisted of 2,005 college students (34.5% men) who answered 40 items assessing four mental health symptoms (i.e., alcohol-related problems, anxiety, hostility, and depression) at baseline immediately before being randomly allocated to a brief alcohol intervention or control group (pre-test) and at a 12-month follow-up following the intervention (post-test). Participants in the intervention group received personalized feedback on their alcohol use and alcohol-related problems, along with descriptive drinking norms of peers and other personalized and general information aimed at motivating students to change. Results indicated that the selected models showed adequate fit and classification outcomes. Latent logistic regression analysis showed that the intervention helped improve participants’ anxiety and depression. Those in the intervention group were more likely to transition from having anxiety and depression attribute profiles at pre-test to not having them at post-test. In addition, male students were more likely to improve anxiety. Although the intervention was not associated with the transition probability from presence to absence for alcohol-related problems, it helped suppress the transition to having the attributes of alcohol-related problems (among men) and hostility (among women) at post-test. However, male students in the intervention were more likely to transition from absence to presence in their depression attribute profile state. The three-step latent transition CDM with covariates showcased in the current study may be an appealing analytical tool for examining and explaining change in mental health symptoms with informative covariates.