Determining perceived meaning during impression formation: another look at the meaning change hypothesis

The meaning change hypothesis says that the individual items of information from which an impression of someone is formed tend to be perceived in a way consistent with the overall impression. It is argued here that the evidence usually considered as the principal experimental support for this hypothesis is artifactual, the product of a spurious method of determining perceived meaning. We describe a new method in which, following a series of person descriptions, subjects try to recall from each description a critical trait with the aid of a roughly synonymous cue word that is evaluatively either consistent or inconsistent with the description as a whole. The meaning change hypothesis predicts a higher level of recall when the cue word and the overall description are evaluatively consistent. We show that although this prediction holds with at least one kind of person description (Experiments 1 and 2), it does not hold with the kind most often used in impression formation research (Experiments 3, 4, and 5).