Expanding and retracting from the self: Gains and costs in switching self-associations

We report 2 experiments to assess the strength of forming and breaking associations to the self, familiar others, and unfamiliar others in a simple shape–label matching task. In each experiment, participants first formed shape–person associations (e.g., triangle-self). Subsequently, they had to relearn the associations with the shapes and labels rearranged (self→stranger in Experiment 1; self→friend in Experiment 2) and they carried out a matching task in which they judged whether shape–label stimuli were as newly instructed or re-paired. There were faster responses and fewer errors on match trials for newly formed self-associated stimuli. In contrast, after switching, reaction times were slower and accuracy was reduced on mismatch trials involving shapes previously associated with the self. The strength of the self-advantage in forming the new association on match trials correlated with the difficulty in switching from the old self-associated shape on mismatch trials. The results indicate that self-reference enhances the binding of associations in memory; this facilitates associations to new stimuli, but there is a cost of interference from old associations.