MEDIÁCIÓS ELEMZÉSEK PSZICHOLÓGIAI KUTATÁSOKBAN

Abstract

Background and Aims: The study of relationships between variables is of particular importance in psychological research. Statistics is available for such analyses through correlation and regression. If more than two variables (X, Y,…) are considered, a prediction problem X → Y may ask whether other variables have a significant mediating effect in the X → Y effect. In this context, this article describes the concept of mediation analysis, its statistical background, its simplest models, and a simple way to perform and represent it using three free softwares:ROP-R, JASP, and jamovi. A number of illustrative examples taken from real psychological research help to understand the essence of mediation analysis.