Relative Association, Interactiveness, and the Bizarre Imagery Effect

Abstract

The memory advantage for bizarre images over common images was investigated as a function of two stimulus conditions: (a) the relative association between to-be-remembered (TBR) items, and (b) the relationship between the action and TBR items, or the interactiveness. In Experiment 1, the relative association between the TBR items was manipulated. Free-recall performance indicated a significant bizarre imagery effect for strongly related TBR items, but a commonness effect for weakly associated TBR items. Experiments 2, 3, and 4 investigated differences in memory recall of bizarre images as a function of the type of immediate cued-recall test used (subject; predicate) and the method of constructing bizarre sentence stimuli (noun order change method; predicate change method). Predicate cued-recall performance indicated a significant bizarre imagery effect for both stimulus construction methods (Exps. 2 and 4). However, subject cued-recall performance did not yield a significant bizarre imagery effect for the noun order change method (Exp. 3). The results indicate that the bizarre imagery effect can be manipulated as a function of the level of association and interactiveness between the TBR items. In addition to defining stimulus conditions necessary for the bizarre imagery effect, the results suggest that a relational mechanism underlies the bizarre imagery effect.