Paying Attention to Emotional Images With Impact F. C. Murphy, E. L. Hill, C. Ramponi, A. J. Calder, and P. J. Barnard Medical Research Council Cognition and Brain Sciences Unit, Cambridge, England Emotional stimuli receive high processing priority in attention and memory. This processing “advantage” is generally thought to be predominantly mediated by arousal. However, recent data suggest that ratings of an image’s affective “impact” may be a better predictor of recollection than arousal or valence. One interpretation of these findings is that high-impact images may draw an individual’s attention, thus facilitating subsequent processing. We investigated the allocation of visual attention to negative emo- tional images that differed in impact but were matched for valence, arousal, and other characteristics. Participants viewed a central image flanked by 2 neutral indoor or outdoor scenes and made speeded judgments about whether the neutral scenes matched. In Experiment 1, responses were slower on high-impact relative to low-impact or neutral trials. In Experiment 2, responses on high-arousal relative to low-arousal trials did not differ significantly. These data provide evidence for differential allocation of attention to distinct sets of negative, equally arousing images, and argue against the prevailing view that heightened attention to and processing of emotional stimuli relate simply to arousal or valence. Keywords: attention, emotion, impact, arousal At any moment in time, the world is filled with sights, sounds, and smells that simultaneously compete for our attention. The human mind is limited in its ability to process this information. Thus, an important role for an efficient attentional system is to select relevant information in the environment that will receive priority for subsequent processing (Desimone & Duncan, 1995). But what is it that makes certain stimuli more relevant, or salient, than others? One characteristic that has received a significant amount of empirical interest in this regard is the affective, or emotional, quality of a stimulus. Converging evidence from studies that use a range of experi- mental paradigms and stimulus types indicates that people pay greater attention to emotional than to neutral stimuli. For example, threatening material has been found to influence attention in the emotional Stroop task, where participants are asked to name the color of an emotional word or picture (Kunde & Mauer, 2008; Williams, Mathews, & MacLeod, 1996). Other paradigms that have been used to study attention to emotional stimuli include the dot-probe paradigm (Bradley et al., 1997), visual search (Öhman, Flykt, & Esteves, 2001), attentional orienting (Fox, Russo, Bowles, & Dutton, 2001; Posner, 1980), and the attentional blink (Anderson, 2005; Barnard, Ramponi, Battye, & Mackintosh, 2005). Together, the findings indicate that emotional information receives privileged access to attention and awareness in both space and time. These interference effects are often heightened in clin- ically anxious participants (Mathews & Mackintosh, 1998; Mathews & MacLeod, 1994) and persist when awareness of threat stimuli is reduced or eliminated by backward masking (Fox, 1996; Mogg & Bradley, 1998). It is generally agreed that emotional stimuli receive preferential allocation of attentional resources because of their special adaptive significance. Threatening facial expressions (Öhman, Lundqvist, & Esteves, 2001) or threatening animals such as snakes and spiders (Öhman, Flykt, & Esteves, 2001) may be thought of as being evolutionarily important. More recent research has tried to address whether such biases are restricted to negative and threat-related stimuli, or whether they generalize to positive emotional stimuli as well. A cognitive system that efficiently detects an appetitive or aversive stimulus and selects the appropriate approach or with- drawal behavior would also contribute to an organism’s survival and well-being (Calvo & Lang, 2004). Pleasant stimuli have shown parallel attentional effects in several recent studies. Ander- son (2005) demonstrated sparing of the attentional blink for neg- ative and positive words. This sparing was not due to nonemo- tional stimulus characteristics that were associated with distinctiveness; Anderson thus concluded that it was emotional arousal, and not valence, that was critical in influencing attentional performance. An eye-tracking study similarly concluded that when pleasant and unpleasant emotional images are matched for arousal, relative to neutral images, they capture visual attention to a similar degree (Nummenmaa, Hyöna ¨, & Calvo, 2006). Such findings have led to the current dominant theoretical account of emotional biases in attention—that these effects are predominantly mediated by emotional arousal. However, an al- most exclusive focus on valence and arousal inevitably draws attention away from other variables that could influence attention to emotional stimuli. Unpleasant and highly arousing stimuli may vary from neutral stimuli on variables that are distinct from the dimensions of pleasantness and arousal. In particular, studies of F. C. Murphy, E. L. Hill, C. Ramponi, A. J. Calder, and P. J. Barnard, Medical Research Council Cognition and Brain Sciences Unit, Cambridge, England. This research was funded by the United Kingdom Medical Research Council under Project Codes U.1055.02.001.00001.01 (A. J. Calder) and U.1055.02.003.00001.03 (P. J. Barnard). Correspondence concerning this article should be addressed to F. C. Murphy, 15 Chaucer Road, Cambridge, United Kingdom, CB2 7EF. E-mail: fionnuala.murphy@mrc-cbu.cam.ac.uk Emotion © 2010 American Psychological Association 2010, Vol. 10, No. 5, 605– 614 1528-3542/10/$12.00 DOI: 10.1037/a0019681 605