False contingency beliefs reverse contingency learning effects in the valence contingency learning task

In the valence contingency learning (VCL) task, participants evaluate target words which are preceded by nonwords, which are predictive for positive/negative evaluation responses. This produces robust contingency learning (CL) effects, reflected in faster and more accurate performance for highly contingent nonword-valence pairings. Previous findings indicate that controlling for episodic retrieval of transient stimulus-response episodes reduces CL effects but does not eliminate them, as a residual CL effect remains. These residual CL effects are best explained by propositional learning. To substantiate this, the present study manipulated participants beliefs about contingencies in the VCL task. Participants received either true, false or no instructions regarding the actual nonword-valence contingencies. Effects of contingency learning and evaluative conditioning (EC) for nonwords were assessed. Contingency beliefs modulated contingency learning, as true instructions boosted residual CL effects; false instructions reduced residual CL effects, relative to the no instruction condition. Exploratory analyses revealed a modulatory influence of contingency beliefs on EC effects, which varied solely as a function of (remembered) contingency instruction and were unaffected by experienced contingencies. The present study conceptually replicates findings from colour-word contingency learning in the realm of evaluative learning. Implications for theories on processes underlying contingency learning and evaluative conditioning are discussed.