Comparison Group Identification for Difficult-to-Evaluate Populations: Lessons from Evaluating 1n10 LGBTQ Youth Support Services

Abstract

The evaluation literature on Lesbian, Gay, Bisexual, Transgender, and Queer (LGBTQ) youth programs is largely absent of quantitative studies, likely in part due to the challenge of using experimental or quasi-experimental evaluation designs. This paper proposes the creative use of a national data set to overcome the problem of estimating a counterfactual for this population. In addition to discussing lessons from this approach, we describe the program and its impacts. Evidence suggests that the program under study—1n10, a local support group for LGBTQ youth in the Phoenix, AZ, metropolitan area—draws a relatively more disadvantaged group than the national average but that their levels of suicide ideation are lower. By addressing key threats to the internal validity of evaluation designs, we rule out plausible rival explanations for program impacts.