Norms for emotion-false memory lists.

Abstract

False memory has been a flourishing research area for decades, and recently there has been considerable interest in how emotional content affects it. Literature reviews have noted a lack of normed materials that vary in emotional valence and arousal as a factor that contributes to the mixed findings on emotion-false memory effects. We report a pool of normed materials of this sort, the Cornell/Cortland Emotional Lists (CEL). This is a Deese/Roediger/McDermott (DRM) type list pool in which words' mean valence and arousal ratings are factorially manipulated across 32 lists. These lists' levels of mean backward associative strength (MBAS) are all high enough to induce significant levels of false memory. The lists were normed by administering them to 228 subjects at three different universities, all of whom responded to recall and recognition tests for the lists. The norming data revealed that false recall and false recognition were higher for negative lists than for positive lists, whereas true recall and true recognition were higher for positive lists than for negative lists. In addition, high arousal strengthened the valence effects on both true and false recall. These results indicate that the CEL lists are useful tozols for emotion-false memory research.