A reevaluation of the evidence for spontaneous trait inferences

In Experiments 1 and 2 subjects under memory instructions studied a list of trait-implying sentences. Recall of sentences was measured under trait cue and no cue conditions, with the cue variable manipulated within subjects in Experiment 1 and between subjects in Experiment 2. The typical cued recall advantage was found only in Experiment 1 and was attributed to differential output interference rather than the spontaneous encoding of trait inferences. In Experiment 3, the cue variable was manipulated between subjects and the recall period was limited to 1 min. Recall under the cue condition was significantly lower than under the no cue condition. These data suggest that trail inferences are not spontaneously encoded. (C) 1996 Academic Press, Inc.