Abstract
Three experiments tested the effect of an attitude towards an object on the memory judgement of whether this object co-occurred with positive versus negative stimuli. We induced positive or negative attitudes towards novel male stimuli, and paired each man with an equal number of positive and negative animals. In a memory test, participants reported more co-occurrences of same-valence man/animal pairs than opposite-valence pairs. This valence-compatibility effect occurred even when attitudes were induced after the pairing (Experiment 1), when participants knew that each man occurred with an equal number of positive and negative animals (Experiment 2), and in reports of clear memory of pairs that did not co-occur (Experiment 3). The present findings suggest that evaluation causes illusory correlation even when the co-occurring stimuli are not traits or behaviours attributed to the attitude object. The results question the validity of co-occurrence memory judgements as measures of co-occurrence awareness in evaluative conditioning (EC) research.
Notes
1 A common misconception is that illusory correlation refers only to the tendency to attribute non-common attributes to non-common groups (see Hamilton & Rose, Citation1980, for an example that used this term for the congruency bias effect discussed in the present article).
2 As could be expected, the results of the analysis of the No response rates mirrored the analysis of the Yes responses, with significant compatibility effects only for clear-memory responses.