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.