A negativity bias in evaluative counter-conditioning

So far, there is little evidence of a negativity bias in evaluative conditioning. In evaluative conditioning, neutral stimuli (conditioned stimuli; CSs) are paired with stimuli of either positive or negative valence; as a result, the initially neutral stimuli change their valence in the direction of the paired valent stimuli. We investigated if a negativity bias occurs when a CS is paired with both negative and positive stimuli, in sequence, in an evaluative counter-conditioning procedure. In three experiments (N = 100, N = 362, N = 120), conditioned stimuli (CSs) were paired with either positive or negative stimuli in an evaluative conditioning phase; then in an evaluative counter-conditioning phase, the same CSs were paired with stimuli of the opposite valence. We tested whether counter-conditioning is more effective when positively conditioned CSs are negatively counter-conditioned than when negatively conditioned CSs are positively counter-conditioned. We found this to be the case. There was no evidence that this negativity bias was driven by differences in memory. Furthermore, we found no evidence that a negativity bias (nor a positivity bias) occurs in a typical (initial) evaluative conditioning procedure. We discuss implications for understanding evaluative conditioning and negativity biases.