The effect of irrelevant pairings on evaluative responses

Abstract

Pairing a neutral object with a valenced stimulus often results in the former acquiring the valence of the latter (i.e., the Evaluative Conditioning [EC] effect). However, the pairing of an object with an affective stimulus is not always indicative of valence similarity. Three preregistered experiments (total N = 1052) explored EC effects when people were explicitly informed that pairings do not reflect valence similarity. In Experiment 1, informing participants that the paired stimuli are unrelated and therefore irrelevant to each other, reduced but did not eliminate EC effects. In Experiment 2, exposure to pairings defined as irrelevant still produced an EC effect, even when participants were asked to resist being influenced by the pairings. In Experiment 3, irrelevant pairings still produced an EC effect, even when alternative diagnostic evaluative information was provided. The results constrain existing theoretical models of EC and suggest that EC effects are robust.