Acceso usuarios
Markus Brauer
The issue of mediation in the study of psychological phenomena is discussed. One speaks of mediation when the effect of an independent variable on a dependent variable exerts itself via a third variable such that the independent variable affects the mediator variable, and the mediator variable affects the dependent variable. After having distinguished mediation and moderation, we present all conditions that have to be satisfied before one can conclude that the data are consistent with a mediational causal model. Then, we discuss particular cases of mediation such as partial mediation, the identification of a mediator variable among several potential candidates, the determination of the causal direction between two dependent variables, the identification of mediation in the presence of moderation, as well as mediation in experimental designs with repeated measures. The conclusion underscores the importance of identifying mediating processes for basic and applied research in psychology.
