FACTORS CONTROLLING THE ADVANTAGE OF SELF-REFERENCE OVER OTHER-REFERENCE

Keenan and Baillet (1980) showed that the self-reference effect (SRE) occurs for evaluative questions, but not for factual questions. Two experiments comparing self-reference to other-reference are reported in an attempt to clarify and extend that result. First, it is shown that even when factual questions tap autobiographical information, there is still no SRE. This result shows that the important factor in determining the SRE is not only the amount of information available in memory for self and others, but also the amount that is actually used in the encoding. Second, it is shown that the finding of SREs for evaluative questions but not for factual questions is not due to the fact that evaluative questions have used trait adjectives while factual questions have used common nouns (Maki & McCaul, 1985). The present experiments used the same nouns for both evaluative and factual questions and found the SRE for evaluative but not for factual questions. It is proposed that evaluative questions produce an SRE because they exploit the imbalance in amount of information stored with the self-schema versus other person schemas, producing more elaborate encodings for self-referenced events relative to other-referenced events. Factual questions do not produce a SRE because they tap comparable amounts of information for self and others, resulting in equivalently elaborated memory traces.