Session: Applying Heirarchical Linear Modeling and SEM to Address Racial and Economic Inequality (Society for Social Work and Research 24th Annual Conference - Reducing Racial and Economic Inequality)

307 Applying Heirarchical Linear Modeling and SEM to Address Racial and Economic Inequality

Sunday, January 19, 2020: 9:45 AM-11:15 AM
Mint, ML 4 (Marriott Marquis Washington DC)
Cluster: Research Design and Measurement (RD&M)
Keith Chan, PhD, State University of New York at Albany, Thanh Tran, PhD, and Kaipeng Wang, Ph.D., Texas State University
Biological evidence insufficiently account for differences in health and economic outcomes for vulnerable populations in the United States. The fabric of the United States is rapidly changing due to demographic shifts, and we are becoming more and more diverse in our race, immigrant, gender and cultural identities. Despite advances in science education, and medicine, there continues to be disparities in outcomes for vulnerable populations, and much of the evidence points to influences that deleteriously impact racial minorities and marginalized groups. Many of these influences which are based in the social and physical environment can lead to disadvantages in the lifecourse and have a community-level impact. Cross-cultural research which utilizes HML and SEM techniques can serve as a powerful tool to address racial and economic inequality, by informing policy decision-making and providing a framework for interventions.

Mediation analysis in SEM is important for addressing racial and economic inequality for several reasons. Past evidence suggest that there are disparities in health and economic outcomes among peoples of different racial, ethic, linguistic, or national backgrounds. The emergence and reinforcements of these disparities operate along different trajectories and pathways for marginalized populations in the lifecourse. Identifying specific pathways for different cultures can improve outcomes through timely and appropriate medical, psychological and social interventions.

Multilevel analysis, or HLM, has an intuitive fit to cross-cultural research. This approach takes into account the heterogeneity of different populations. Cross-cultural multilevel analysis can account for “neighborhood effects,” which can explain racial and economic inequality at mezzo and macro levels. At the same time, multilevel analysis can be integrated with mediation analysis to apply further rigor and explanatory power to applied cross-cultural research. This can allow social work researchers to examine and address racial and economic inequality in the trajectories of vulnerable populations in individual and community levels.

This workshop aims to teach participants HLM and SEM approaches which can be employed in cross-cultural social work research. The workshop will provide participants with step-by-step illustrations in the use of SEM and multilevel approaches through hands-on exercises, to compare cross-cultural population using large-scale, population-based survey data. These techniques have important applications in social work research, especially in providing a framework of evidence to examine health disparities in empirical, cross-cultural research. For each statistical approach discussed in this workshop, we will explain the underlying purpose, basic assumptions, types of variables, application of commonly used statistical packages (e.g., SPSS, Stata, Lisrel, HLM), the presentation of statistical findings, and the interpretation of results, as well as how to explain the implications of these findings for policy and practice.

After attending this workshop, participants will be able to 1. Explain the foundation of cross-cultural research, from the definitions of culture in research, the importance of understanding risk and protective factors, and conceptualization of social problems across vulnerable populations. 2. Identify statistical approaches and techniques relevant to examining similarities and differences in cross-cultural research 3. Apply the appropriate use of mediation and multilevel analysis to tease out differences in effects and causal pathways in cross-cultural research

See more of: Workshops