Qualitative/Quantitative Synergies in a Random-Assignment Program Evaluation

Greg J. Duncan, Northwestern University

This paper describes the productive synergy that resulted from a mixed qualitative/quantitative approach to evaluating program impacts of New Hope, an anti-poverty program developed and implemented in Milwaukee, Wisconsin in the mid-1990s. An overarching conclusion is that the combination of the two methodologies can indeed enrich evaluation efforts. The assessment of New Hope was enhanced by decisions to randomly sample qualitative cases from the larger population of New Hope participants and to train graduate student research assistants to both conduct the qualitative interviews and analyze the quantitative data. The mixed method approach was helpful for understanding program impacts estimated in the quantitative data and for identifying subgroups for which program impacts were the strongest. The n=43 qualitative data were helpful for identifying some but not all kinds of experimental impacts. Case studies from the qualitative data provided important context for the evaluation report and were also used to generate measures for a follow-up survey.