Original Article Venom, speed, and caution: effects on performance in a visual search task Danielle Sulikowski a,b, ⁎ a School of Psychology, Charles Sturt University, Bathurst, NSW, 2795, Australia b Department of Brain, Behaviour and Evolution, Macquarie University, Sydney, NSW, 2109, Australia Initial receipt 30 May 2011; final revision received 15 November 2011 Abstract Previous reports of faster responses to threatening compared to benign stimuli in visual search tasks have argued that threatening targets are faster to engage and slower to disengage attention than benign targets. This study reinterprets previous findings and resolves inconsistencies in the literature by replacing the theory of differential disengagement of attention with one of differential caution. It also examines whether visual attentional mechanisms are sensitive to more than just the threatening versus benign categorical status of the targets and introduces a novel measure (a caution score) that appears to be sensitive to the level of threat implied by the target image, but immune to other stimulus features (target-distracter similarity and threat status of distracters) known to affect reaction time. As well as locating threatening targets faster than benign targets, participants were also faster, more accurate, and more cautious to detect lethal spiders compared to nonlethal spiders and even more cautious again if the spiders were presented on a person's hand. These results suggest that mechanisms of attention and threat evaluation interact during visual search tasks, producing behaviour that is sensitive to the target's implied threat level and the context in which that target is presented. © 2012 Elsevier Inc. All rights reserved. Keywords: Visual search tasks; Threat detection; Attention; Caution 1. Introduction Visual search tasks have been used extensively to assess how ecologically relevant stimuli capture attention, reveal- ing faster and sometimes more accurate identification of threatening stimuli, such as angry faces (using both photographs, Gerritsen, Frischen, Blake, Smilek, & East- wood, 2008; Hansen & Hansen 1988; and schematics, Fox et al., 2000; Öhman, Lundqvist, & Esteves, 2001) and snakes and spiders (first demonstrated by Öhman, Flykt, & Esteves, 2001, and subsequently affirmed by others, including Blanchette 2006; Brosch & Sharma 2005; Flykt 2005; and Fox, Griggs, & Mouchlianitis, 2007), when compared to nonthreatening stimuli. Both children (LoBue & DeLoache 2008) and prelingual infants (LoBue & DeLoache 2010) respond to snakes more rapidly than flowers, suggesting that an initial perceptual bias is present early in development and may be independent of experience or explicit knowledge of danger. LoBue (2010) has also demonstrated, however, that in children recent aversive experience with potentially dangerous man-made objects (syringes) is associated with those objects being located faster. Taken together, these findings suggest that preexisting biases interact with experiences that induce fear to shape the mechanisms responsible for biasing attention toward potentially threat- ening objects. Typically, these tasks use images from the benign target category as distracters for the threatening targets and vice versa (Brosch & Sharma 2005; Öhman, Flykt, et al., 2001). For the threatening targets, a target-present trial would present one image of the threatening target amongst several images of the benign distracter, while a target-absent trial would only present images of the benign distracters. For the nonthreatening targets, this arrangement would be reversed with the threatening images now being presented as the distracters. Participants are typically required to respond by declaring all items to be the ‘same’ (meaning the target is absent) or that one item is ‘different’ (meaning the target is present; Fig. 1 presents two versions of this typical paradigm and the modified paradigm of the present study). Although successful in identifying differences in response times Evolution and Human Behavior 33 (2012) 365 – 377 ⁎ Corresponding author. School of Psychology, Charles Sturt Univer- sity, Bathurst, NSW, 2795, Australia. E-mail address: danielle.sulikowski@ymail.com. 1090-5138/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.evolhumbehav.2011.11.007