From Deviant Likes to Reversed Effects: Re-Investigating the Contribution of Unaware Evaluative Conditioning to Attitude Formation

Evaluative Conditioning (EC) is the change in liking of an object due to its mere pairing with a positive/negative stimulus. A central question in EC research is whether EC effects also emerge without awareness of the stimulus pairings. This is often tested by asking participants after the conditioning whether an object had been paired with positive or negative stimuli. If participants’ answers in these memory measures mismatch with the US valence (e.g., “positive” response when an object was paired with a negative stimulus), the pairings are classified as unaware. The last decade of EC research has found mostly no evidence for unaware EC, and sometimes even reversed unaware EC effects when using such memory measures. The present work demonstrates that such valence memory measures underestimate unaware EC effects due to differences between the normed and the subjective US valence. In two simulation studies, a re-analysis of previous studies, and four preregistered experiments (N = 502), we assess when this bias is more or less severe, depending on common procedural variations in EC experiments. We also propose an improved memory measure of aware/unaware EC. Yet, even when the bias was reduced in the corrected measure, no evidence for unaware EC could be found. Overall, our research shows that unaware EC may be difficult to detect with valence memory measures. Also, they support current memory-based and propositional EC accounts.