Addressing Intervention Fit When Developing Services for Adolescents Leaving Care - An Empirical Assessment of Measurement Invariance and Latent Mean Differences in Theoretically Defined Mediators.

Abstract

Purpose: A good fit between program theory and the needs of the target population is essential to the success of an intervention. Still, empirical investigation of fit is a rare activity in theoretical frameworks for intervention development. The study objective was to inform the development of relevant support for youth transitioning from out-of-home care (OHC) in Sweden in light of intervention fit. Self-efficacy, mental health, resilience, and self-determination have been identified as potentially important malleable mediators that can be targeted by independent living services (ILS). This study aimed to establish the extent to which measurement instruments used to capture these mediators measure similarly across youth that belong to a potential target population for ILS and those who do not.

Materials and methods: A cross-sectional survey design was used to compare youth with and without experience of OHC on the four study outcomes using a multi-group measurement invariance analysis and latent mean comparison.

Results: All scales demonstrated scalar invariance or partial scalar invariance. Contrary to initial hypothesis, no differences in latent means were found.

Discussion: Four potential interpretations of the results were discussed. Conclusion: The study identified a potential discrepancy between theoretical assumptions and empirical characteristics relevant to development of ILS for youth with experience of OHC in Sweden. Moreover, it highlighted the importance of exploring theoretical core mediators empirically in the target population in the process of intervention development and evaluation.