Social Cognition Network Meeting edition:2 location:Heidelberg date:31 August - 3 September 2000
The relationship between the degree to which people ascribe a trait to a target and the amount of trait-relevant behavioural information they can remember about that target is sometimes surprisingly weak. According to Hastie and Park (1986), this is because people create on-line personality impressions which are stored in memory separately from the information on which they were originally based. When asked to report their impression, people retrieve their impression without referring to behavioural information. We propose that trait ratings are derived from retrospective act frequency estimates that are sometimes independent from the recall of behavioural instances so that trait ratings and behaviour recall may be relatively independent.
To compare both views, participants received a list of thirty behaviours (10 extroverted, 10 introverted and 10 neutral) either with an impression formation or with a memory instruction set. A biased recognition test was administered to manipulate retrospective act frequency estimates in an extroverted or an introverted direction, presumably after participants had created their on-line impressions. Participants were asked ratings of the target’s extroversion and retrospective act frequency estimates. They were also administered a recall test concerning the original information. Supporting our view, all measures were influenced by the biased recognition test.