Emotional recognition memory for younger and older adults: Combining ROC analysis and the diffusion model
Three experiments investigated the effect of emotional stimuli on recognition accuracy and response bias for younger and older adults using ROC analysis and Ratcliff's (1978) diffusion model. Theoretically, emotion may enhance memory accuracy either by improving encoding processes or by altering the memory consolidation process. These competing hypotheses were evaluated in a recognition experiment that tested memory both before (immediate testing) and after (20 minute delay) the consolidation process would likely be completed. The emotion-specific consolidation hypothesis was not supported: there was no interaction of emotional-valence with test delay. Because previous research has shown that negatively-valenced items consistently lead to more liberal responding for both older and younger adults, and inconsistently affect memory accuracy (Kapucu, Rotello, Ready, & Seidl, 2008), confidence ratings and reaction time data were assessed. These data were modeled with signal-detection and diffusion approaches that allow independent measurement of memory accuracy and response bias effects. Although the two methods did not converge for all subjects, in general negative words led to large shifts in response bias and increased recognition accuracy for both younger and older adults.