A latent profile analysis of at-risk students in a multilevel prevention program: Who chooses to participate and the impact on outcomes

Abstract

The goal of the current study was to adopt a novel comprehensive approach to examine the efficacy of the targeted intervention component of a multilevel prevention program. We addressed four fundamental issues with this approach: (1) Distinct subpopulations within the sample; (2) Characteristics of those who chose to complete the intervention and those who did not by profile; (3) Changes in the experimental group by profile; and (4) Changes in the comparison group by profile? Latent Profile Analyses revealed that our initial sample of students identified as at-risk due to current symptoms of distress, clustered into three distinct profiles of risk (High Risk (HR), Average Risk (AR) and Low Risk (LR)). Regarding characteristics related to opting for active participation, we found that the higher the level of initial risk, the more likely students were to opt to participate in the program, and within the AR profile, we found that those participants with lower study skills opted more often for the intervention. Latent Growth Modeling, in the experimental group, showed that the HR profile benefited substantially from participation in the intervention with significant reductions in symptoms and risk factors, as well as significant increases in protective factors. Changes in the AR profile were in the expected direction, but this group remained notably anxious although not depressed anymore. The LR profile showed significant reductions in symptoms following the intervention. In the comparison group, the HR profile improved some, the AR and the LR did not change significantly, except for significant decreases in personal goal and values consolidation in the LR profile. The current study highlights the importance of adopting a more comprehensive evaluation of efficacy and how this must inform selection criteria for future studies.