False Memory in Aging: Effects of Emotional Valence on Word Recognition Accuracy Olivier Piguet Massachusetts Institute of Technology and Prince of Wales Medical Research Institute Emily Connally, Anne C. Krendl, Jessica R. Huot, and Suzanne Corkin Massachusetts Institute of Technology Memory is susceptible to distortions. Valence and increasing age are variables known to affect memory accuracy and may increase false alarm production. Interaction between these variables and their impact on false memory was investigated in 36 young (18 –28 years) and 36 older (61– 83 years) healthy adults. At study, participants viewed lists of neutral words orthographically related to negative, neutral, or positive critical lures (not presented). Memory for these words was subse- quently tested with a remember– know procedure. At test, items included the words seen at study and their associated critical lures, as well as sets of orthographically related neutral words not seen at study and their associated unstudied lures. Positive valence was shown to have two opposite effects on older adults’ discrimination of the lures: It improved correct rejection of unstudied lures but increased false memory for critical lures (i.e., lures associated with words studied previously). Thus, increased salience triggered by positive valence may disrupt memory accuracy in older adults when discriminating among similar events. These findings likely reflect a source memory deficit due to decreased efficiency in cognitive control processes with aging. Keywords: emotion, learning, memory deficit, response bias Memory is susceptible to distortions. In real life, individuals claim to remember events that never happened or misremember portions of events that did take place. In the laboratory, false memory occurrence has been studied in experiments that use lists of words phonetically or semantically associated with specific target words (i.e., Deese–Roediger–McDermott, or DRM, task; Deese, 1959; Roediger & McDermott, 1995). At study, partici- pants view or hear a list of words that are related to a target word (lure) that is not presented. Subsequent recognition memory testing of the studied words consistently gives rise to false recognition of the lure despite the fact that this word was never studied. Production of false alarms is modulated by the emotional content of the stimuli. Pesta, Murphy, and Sanders (2001) studied false recognition of emotional and neutral words using the DRM paradigm. Their results demonstrated that the rate of false alarms varied according to the distinctiveness of the emotional lures: Participants were more likely to endorse emo- tional lures when the study list included other emotional words than when it did not. Regardless of the distinctiveness, how- ever, false alarms to emotional lures were lower than to neutral lures, supporting an independent effect of emotion on memory accuracy. In contrast, Windmann and Kutas (2001) reported more false alarms to emotional than to neutral words during a recognition memory task. They proposed that emotion induced an attention bias, which resulted in an increased likelihood of recognizing an emotional stimulus, regardless of prior exposure to it. Healthy aging is accompanied by significant changes in ep- isodic memory. Overall, the age-related decline is particularly pronounced for correct responses that indicate a clear and vivid recollection of previously presented items (i.e., remember re- sponses; Tulving, 1989) and less so for responses reflecting familiarity with the item in the absence of specific details associated with the experimental context (i.e., know responses). Aging also sees an increase in incorrect or inaccurate responses (Roediger & Geraci, 2007). During list-learning tasks, older adults are not as effective as young adults in differentiating between studied and unstudied items and are more likely to endorse an item as “old” rather than reject it (Bastin & Van der Linden, 2003; Jacoby, Bishara, Hessels, & Toth, 2005). Fur- thermore, manipulations of list presentations in which DRM paradigms, compound words, or recombined words are used consistently reveal increased false alarms in older adults com- pared with young adults (Jones & Jacoby, 2005; Reinitz & Hannigan, 2004; Watson, McDermott, & Balota, 2004). These results suggest that incorrect responses may be retrieval errors due to reduced monitoring systems and decreased efficiency in Olivier Piguet, Department of Brain and Cognitive Sciences, Massachu- setts Institute of Technology; and Prince of Wales Medical Research Institute, University of New South Wales, Sydney, Australia; Emily Con- nally, Anne C. Krendl, Jessica R. Huot, and Suzanne Corkin, Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology. Emily Connally is now at the University of Arizona, Tucson. Anne C. Krendl is now at Tufts University. Some of these results were presented at the 35th Annual Conference of the Society for Neuroscience, Washington, DC, November 2005. This research was supported by National Institutes of Health Grant AG021525. Olivier Piguet was supported by National Health and Medical Research Council of Australia Neil Hamilton Fairley Postdoctoral Fellowship 222909. Correspondence regarding this article should be addressed to Olivier Piguet, Prince of Wales Medical Research Institute, Barker St., Randwick NSW 2031, Australia. E-mail: o.piguet@powmri.edu.au Psychology and Aging Copyright 2008 by the American Psychological Association 2008, Vol. 23, No. 2, 307–314 0882-7974/08/$12.00 DOI: 10.1037/0882-7974.23.2.307 307