The Valence Illusion

Abstract

The valence illusion hypothesis is developed and states that valence acts as superordinate dimension in impression formation. Valence biases the judgment of perceived characteristics of other human beings. Evaluative Conditioning (EC) is a procedure to isolate and manipulate valence directly and was used in seven experiments to produce desired valence effects. Three experiments were devoted to the question whether EC procedures can account for valence effects in impression formation. Photographs of persons were conditioned with either positive or negative valence. Altering the photographed persons slightly (e.g., adding glasses) after the conditioning phase simulated everyday change in appearance of human beings. Four additional experiments investigated the proposed role of valence in impression formation. In the evaluative rating phase, participants were asked to rate the portrayed persons on a total of ten different characteristics, all of them belonging to the big two of impression formation, agency and communion. Evaluative Conditioning affected valence ratings on altered photographs, and valence did influence the perceived characteristics of photographed human individuals on all tested characteristics. Furthermore, the size of the valence effect can be predicted by the conceptual similarity of one characteristics to likability. Implications and avenues of future research are discussed.