Beyond Difference Scores: Unlocking Insights with Polynomial Regression in Studies on the Effects of...

The aim of our study was to investigate whether theories of congruence are better tested using polynomial regression analysis, rather than expressing discrepancy between implicit and explicit measures as continuous or categorical difference scores. This paper also aims to make knowledge more accessible by providing a step-by-step explanation of both methods, illustrating differences between them, and making materials openly available for other researchers. In this paper, implicit and explicit measures of self-esteem are used as predictors for depressive symptoms, anxiety, and aggression in a general population sample (N = 135). Explicit self-esteem was measured using the Rosenberg Self-Esteem Scale, implicit self-esteem was measured using the Implicit Association Test, and the Symptom Questionnaire was used to measure depressive symptoms, anxiety, and aggression. The results show those difference score models all imply that the discrepancy between implicit and explicit self-esteem explains depression and anxiety, but not aggression. However, polynomial regression analysis shows that depression and anxiety are not accounted for by the explicit-implicit discrepancy as such, but are foremost explained by explicit self-esteem. Polynomial regression has the potential to evaluate more complex and more detailed hypotheses than what would be possible using statistical approaches based on discrepancy scores. It is therefore recommended for future research aimed at disentangling the roles of explicit and implicit self-esteem in psychological outcomes.