Qualitative Mediation Analysis: an Important Method for Exploring Mediating Mechanisms in Prevention Science.

Mediating variables serve a primary role in devising intervention theories and applying them to practice. Prevention scientists have repeatedly called for understanding how and why an independent variable (X; e.g., intervention) is related to a dependent variable (Y; e.g., drug use). Quantitative mediation is used to describe mediating variables that intervene in the causal path from X to Y. Most methodological development for mediation analysis has focused on statistical methods and the assumptions necessary for valid application of these statistical methods. The current paper describes how qualitative methods extend into mediation research and the unique strength of qualitative mediation in identifying potential mediators and mechanisms of change not previously hypothesized. Taking examples from prevention research, we outline how qualitative mediation generates unique and complementary information about mediating mechanisms that may only be available through interviews, focus groups, observation, archival analysis, and other qualitative methodology. We briefly describe cautions when using qualitative mediation including reliance of retrospective reports, potential to influence interviewees, and selective sampling. We underline how qualitative mediation analysis is particularly well suited for exploratory studies and extracting mechanisms of action for new or adapted interventions in prevention science.