Response vs. Perception

Abstract

Three experiments were conducted in which college students read, and then attempted to match, a series of written descriptive passages with the referent photographs on which they were based; the photographs sho wed the face of an actor, representing a variety of emotional expressions. In Experiment I, subjects provided with a series of context passages depicting a narrow range of emotions (neither pleasant nor unpleasant) chose “matches” having more extreme pleasantness values than did subjects provided with context passages depicting a wide range of descriptions on the pleasantness dimension when responding to test descriptions embedded within the context series. In Experiments II and III, contrast effects were obtained; subjects who had read mostly unpleasant context passages chose more pleasant referents in response to neutral test descriptions than did those who had read mostly pleasant descriptions. The results of all three experiments suggested that these effects were mediated in large part by a response bias, the tendency to use each response alternative with roughly equal frequency. In Experiments II and III, there was suggestive evidence for the possibility that a more central (or perceptual) mechanism may also have contributed to the observed results.