Control variables in marketing research

In empirical marketing research that does not rely on fully randomized experiments, control variables are an important tool to rule out rival alternative explanations for the observed relationships. Despite their importance for causal inference, control variables often receive little attention from either applied researchers or methodologists. At the same time, overviews of control variable practices in neighboring disciplines demonstrate that researchers struggle with selecting, analyzing, and interpreting control variable results. In response, this article combines a synthesis of the theoretical knowledge on control variables with a review of control variable practices. Against this background, we develop and discuss sixteen recommendations for control variable use in marketing research.