Mixed Methods and Causal Analysis

Abstract

Interest in and use of mixed methods research in the social sciences has grown tremendously in recent years and has the potential to assist in addressing core challenges in causal inference. We discuss ways in which the addition of a qualitative component can serve multiple roles in causal analyses, including understanding treatment definition, concept measurement, selection into treatment, causal effect mechanisms, and effect heterogeneity. We also describe how quantitative and qualitative methodologies can be and have been combined in studies seeking to make causal claims, highlighting some of the key research design decisions in integrating qualitative and quantitative methodologies. Given the ability of qualitative research to greatly enhance quantitative studies concerned with causality, we argue that more causal studies should incorporate mixed methods approaches into their research designs.