When reality monitoring fails: The role of imagination in stereotype maintenance.

Research on reality monitoring (the process by which people distinguish memories of real events from memories of imagined events) suggests that the occurrence of imagined events can inflate the perceived frequency of corresponding real events. When the availability of real events is assessed, reality monitoring apparently fails to exclude some events that were only imagined. We conducted two experiments to examine how such failures in reality monitoring can contribute to the mainte-nance of social stereotypes. When subjects imagined members of occupational groups in the initial experiment, they tended to incorporate stereotyped traits into their imaginations, with specific traits determined by the contexts being imagined. This result suggests that imagined events do correspond with stereotype-confirming real events. In the second experiment, subjects read sentences that pre-sented traits (stereotyped and nonstereotyped) in association with occupations with uniform fre-quency. They also imagined members of each occupation in situations relevant to particular stereo-typic traits, but without presentation of the traits. In subsequent judgments of presentation fre-quency, subjects overestimated their exposure to stereotypic occupation-trait combinations, which replicated earlier studies. More important, subjects further overestimated the presentation frequency of imagined stereotypic combinations, which indicated the failure of subjects to distinguish their