What Works in Boston May Not Work in Los Angeles: Understanding Site Differences and Generalizing Effects from One Site to Another
Multi-site interventions are common in numerous fields of research, including education. One example is the Moving to Opportunity (MTO) study, an intervention in which families in public housing were randomized to receive housing vouchers and logistical support to move to low-poverty neighborhoods. This intervention took place in 5 sites across the U.S. and intervention effects varied by site. To date, there has been no quantitative examination of the underlying reasons for these site differences. Our objective is to examine underlying reasons for site differences in the MTO intervention effects on high school dropout among adolescents. Specifically, we wish to test whether differences in effects across sites can be explained by differences in population composition. To do this, we develop and employ a targeted minimum loss-based estimator (TMLE) for the intent-to-treat average treatment effect and complier average treatment effect. We illustrate how a TMLE can incorporate a set of pre-treatment variables and transport formulas to estimate expected effects across sites in the MTO data.
Kara Rudolph is a Robert Wood Johnson Foundation Health and Society Scholar at UC Berkeley and UC San Francisco. She graduated with a PhD in Epidemiology and MHS in Biostatistics from Johns Hopkins School of Public Health in 2014. Her research interests are in applying causal inference methods to studying social and contextual influences on mental health and substance use in disadvantaged, urban areas of the United States.