Finite Mixture Modeling of Community Violence Exposures and Health Related Quality of Life (Pilot Year 7)
David Chen, ChristianaCare (PI); LeRoi Hicks, ChristianaCare (Mentor); Yasser Payne, UD (Co-I)
What we wanted to learn and why it matters
Wilmington, DE has high rates of firearm violence. Over 70% of Wilmington residents witnessed or heard community violence. Such exposures negatively impact mental and physical health. Statistical modelling approaches such as Latent Class Analysis (LCA) are commonly used to identify distinct grouping of multiple related factors that contribute to a given outcome. Engaging community members in the interpretation of LCA modelling results may enhance contextual understanding and practical applicability of statistical results. In this study community, university and health system researchers partnered to conduct and interpret LCA modeling of data collected in the Wilmington Street Participatory Action Research Health Survey (N=782 adults). This study assessed exposures to violence and crime, employment and educational opportunity, and health-related quality of life.
What we did and learned
Five community-based participatory action researchers learned LCA concepts, collaborated on model development, and conducted six focus groups with community members (N=48) to obtain input on visual representations of models and class interpretations. LCA models of community and personal violence exposures were developed. Findings indicated multiple contributing factors to structural violence, beliefs and attitudes, and health-related quality of life. Community inputs further strengthened interpretation and understanding of how identified factors related to structural violence, beliefs and attitudes, and health-related quality of life.
To read more about this study
Chen D, Alleyne A, Tucker S, Payne Y, Hitchens B, Cornish B, Price L, Copeland K, Chamber D, Caplan R, Russell T, Middlebrook N. Community engaged latent class analysis: mixed methods approach to the intersection of community violence exposure and health.