Remembering things that never occurred: the effects of to-be-forgotten stereotypical information

Participants, 68 female and male nonpsychology university students, studied stereotypical and nonstereotypical words either with an initially activated social category (immigrant prime) or with no social category (neutral prime). They were then instructed to either forget or remember the studied words. This was followed by a recognition test. Based on the Deese-Roediger-McDermott (DRM) paradigm and stereotype-accessibility research, we anticipated that the participants in the immigrant, but not in the neutral, prime condition would falsely recognize more nonstudied stereotypic attributes under a forget than a remember instruction. The results supported our hypothesis. Implications of the outcome for eyewitness testimony, juridical decision-making, and stereotype maintenance are discussed.