TEORIJSKI I PRAKTIČNI ZNAČAJ SLOŽENIH MODELA ZASNOVANIH NA ANALIZAMA MODERACIJE I MEDIJACIJE

Abstract

Most statistical methods applied in the social sciences involve testing direct relationships among variables. However, since psychology deals with complex phenomena, this corpus of methods is not sufficient to understand the mechanisms and conditions under which certain direct relationships apply. Moderation and mediation analyzes were the first step towards understanding the more complex relationships among variables. The analysis of mediation answers the question of “how” and “why”, or through what mechanisms a particular relation is realized, while the analysis of moderation offers the answer to the question of “when”, “for whom” and “under what conditions” a particular relation between the predictor and criterion variables applies. Although the increasing use of moderation and mediation analyzes has led to advances in understanding the phenomena addressed by psychology, complex psychological theories require testing more complex models based on these analyzes. Accordingly, this paper will outline theoretical rationales for using models such as moderated mediation, moderation mediation, parallel and serial mediation, three-way interaction, and nonlinear moderation. This paper focuses on explaining the conceptual differences between these complex models, through demonstrating the need to use these models on examples of different psychological theories.