Memorial consequences of variation in behavior perception

Three studies examined the effect of variation in behavior perception on memory for an observed other's behavior. In Study 1, subjects were instructed to segment an actor's behavior into either fine, natural, or gross units of action. Fine-unit subjects subsequently remembered more details about the actor's behavior than did natural-unit subjects who, in turn, remembered more details than did gross-unit subjects. This result provides the first direct evidence for D. Newtson's (1973, Journal of Personality and Social Psychology, 28, 28–38) assertion that fine-unit perceivers are in a higher information state with regard to an observed other than are gross-unit perceivers. In Study 2, fine-unit subjects again significantly outperformed gross-unit subjects in terms of the number of action-related details they could remember even when memory was distorted by misleading postevent information. This result indicates that the link between unitization and action memory is quite robust. Study 2 also assessed subjects' memory for the nonaction-related or peripheral details of a behavior sequence. Reversing the pattern for action-related memory, gross-unit subjects performed better on a nonaction memory test than did fine-unit subjects. Thus, Newtson's (1973) hypothesis that a finer level of behavior perception is associated with greater information gain seems to hold only for action-related information. Finally, in Study 3 some artifactual interpretations of the results of Studies 1 and 2 were assessed and deemed untenable. Implications of the present findings for person memory, eyewitness testimony, and behavior perception research are discussed.