Abstract
Mediation analysis tests whether the effect of an independent variable on an outcome is transmitted through an intervening variable, offering insights into the mechanisms that shape complex relationships. It is widely applied in marketing research, and over time, several approaches have been developed to test mediation (e.g., Baron and Kenny’s method, the bootstrap, and the joint significance test). However, some applications contradict the principles of these methods, undermining both theory building and managerial decision-making. To our knowledge, this is the first systematic literature review of mediation analysis in marketing research. We evaluate current practices, identify common pitfalls, outline steps to avoid them, and assess the strengths and weaknesses of widely used approaches. Drawing on 2209 distinct studies identified within 1,870 articles across 20 leading journals (2013–2023), we employed a structured screening and coding process. Our results provide methodological guidance to improve rigor, transparency, and both theoretical and practical relevance of future mediation analyses in marketing research.