Testing the generality and automaticity of self-reference encoding with release from proactive interference

In three experiments the release from proactive interference paradigm was used to examine the role of the self in memory. In Experiment 1 stimuli were trait adjectives that had been previously rated for their descriptiveness of the rater or for their descriptiveness of a well-known other (Ronald Reagan). Across three trials, significant proactive interference was obtained for all adjectives (regardless of whether they described the self, Ronald Reagan, or neither). Release from proactive interference was observed only for self-rated adjectives. The interpretation of this result was clouded, however, because the trait adjectives also shifted on an evaluative dimension from Trials 1-3 to Trial 4. In Experiments 2 and 3, when the evaluative dimension was not allowed to change, no release from proactive interference was obtained. These results suggest that the involvement of a self-schema in memory is not as ubiquitous or automatic as some views have assumed.