The role of episodic retrieval in evaluative conditioning: evaluative conditioning effects differ depending on the temporal distance to the last stimulus pairing.

Abstract

Liking of a previously neutral stimulus can change based on co-occurrences with positive or negative stimuli, a phenomenon denoted evaluative conditioning (EC). Prior research suggests that EC depends on information about previous stimulus pairings that is remembered when the neutral stimulus is evaluated. According to recent findings from contingency learning, the temporal distance to the last occurrence of a given stimulus determines which information about its previous occurrences is retrieved, favouring recent information after short temporal delays and frequent information after longer delays. In this research, we tested whether EC follows the same retrieval principles. Across three online experiments, we found that EC reflected the valence of the most recently paired valence, but not of the most frequently paired valence, when measured shortly after a pairing. Conversely, when measured after a longer temporal interval, EC reflected the most frequently paired valence, but not the most recently paired valence. These results support a role of episodic memory and retrieval in EC. Our research highlights parallels to contingency learning and suggests that episodic memory processes govern various types of learning resulting from stimulus contingencies.