FAMILY STRENGTHENING: A DEEPER LOOK AT DRIVERS OF PROGRAM SUCCESS

Howard University, Washington DC

Investigators

Abstract

Many studies document large differentials of marriage success by race and age, but the causal predictors, the factors that reliably identify outcomes and point to the factors that contribute to marriage success have yet to be clearly articulated or are contested. In this project, Howard University researchers will study three research questions drawing on three rich longitudinal datasets generated through the Health and Human Services/Administration for Children and Families supported family strengthening programs. Secondary analysis of these databases will provide additional analysis and reporting on impacts of race, ethnicity, location, socio-economic conditions, and age on understanding the success rate of the three programs. The expected outcomes will help agencies craft improved and culturally competent programming aimed at family strengthening. The research questions this study will address are: (1) Did race and ethnicity-neutral outcome measures generate biased estimates of program impacts in the evaluation of the three programs?; (2) Do local social and economic conditions affect the distribution of program impacts?; and (3) Were program interventions more effective on Hispanic participants than on African American and white participants? Hypotheses related to these research questions will be tested through more detailed factor analysis and econometric applications than that done in previous research studies. The goal of this work is to enrich our knowledge of drivers of program success across various sub-populations. Sample: Data samples from the Supporting Healthy Marriage and Building Strong Families programs and the U.S. Census Bureau Measures: The study will estimate regression equations to test the hypotheses. The coefficients from the estimation and their comparisons are the measures used to test the hypotheses and thus provide answer to the three research questions. The study will reconstruct the outcome variables of interest for blacks, whites and Hispanics separately to improve their internal consistency and reliability.