Looking versus seeing: Strategies alter eye movements during visual search

Abstract

Visual search involves the coordination of looking (moving one's gaze to new locations) and seeing (distinguishing targets and nontargets). These two aspects of visual search are distinct from one another because high-acuity vision is possible only in a small region at the center of gaze (the fovea), and only when the eyes are stationary (a fixation). To sample detailed information from an extended scene, the eyes must move abruptly (saccade) from one location to another. In the typical inspection of a scene, this fixation-saccade cycle is repeated 3-4 times/sec.The efficiency of visual search-how rapidly and accurately the target is found-is typically measured by the time that elapses between the first glimpse of a scene and a response indicating target detection. This entails a direct trading relation between seeing and looking: Longer fixations increase information fidelity from each location at the cost of exploring fewer locations, whereas quickly exploring many locations results in reduced fidelity at each one. Studies comparing human oculomotor behavior with an ideal psychophysical observer have indicated that many participants come close to optimizing this trade-off in search (Najemnik & Geisler, 2005.In the present study, we explored the consequences of adopting particular cognitive strategies on this trading relationship. Several studies have shown that participants who are instructed to search passively search more efficiently than those who are instructed to search actively (complete instructions are in the Method section) (Smilek, Dixon, & Merikle, 2006;Smilek, Enns, Eastwood, & Merikle, 2006). Smilek, Enns, et al. (2006) hypothesized that the passive strategy gives automatic processes more influence over spatial attention, whereas the active strategy encourages greater reliance on unnecessary executive processes (cf. Wolfe, Alvarez, & Horowitz, 2000). This interpretation was bolstered by a second experiment in Smilek, Enns, et al. (2006) showing that search was improved when participants performed a simultaneous task that occupied executive processes.In the present study, we asked three broad questions concerning cognitive strategies and eye movements. First, is there any relationship between the two at all? It may be that strategy has no effect on eye movements, in that all of the participants use their eyes to sample information in essentially the same way. If so, the passive advantage found by Smilek, Dixon, and Merikle (2006) and Smilek, Enns, et al. (2006) may be purely cognitive, reflecting differences in the way scene information is processed after the eyes have sampled it.Second, if strategies alter eye movements, which oculomotor measures are affected? We hypothesized that passively instructed searchers will shift their emphasis to looking less and seeing more, spending more time on individual fixations than do active searchers. We also expected differences in other oculomotor behaviors. There are at least two ways one could see more: by expanding the attentional window of each fixation (i...

Person perception informs understanding of cognition during visual search

“…Search performance is summarized by way of an efficiency score, which is obtained by dividing correct RT by the proportion of correct responses (PC) for each participant in each condition. This score is a convenient way to combine search time and accuracy when they are related, as they are here, because it corrects RT values that are underestimated when participants are willing to trade errors for RT (Townsend & Ashby, 1983;Watson et al, 2010). Figure 4 demonstrates that each of these three factors had sizable influences on search efficiency, thus providing the appropriate context for the study of person perception in Phase 2.…”

“…The only factor that was systematically varied between participants was the instruction concerning cognitive strategy (Smilek et al, 2006b, Watson et al, 2010. A random half of the participants were assigned to the active group and were told the following:…”

“…The strategy manipulation involved instructing a random one half of the participants to adopt an active search strategy, whereas the other one half of the participants were instructed to adopt a passive search strategy. This search strategy manipulation has been shown to create a passive advantage in search, whereby passive searchers are more efficient than active searchers (Smilek et al, 2006a(Smilek et al, , 2006bWatson et Fig. 1 Overview of the research design and data analyses, showing the relations between the various measures and how these measures were analyzed and a guide to terminology Fig.…”

“…As such, we already know much about how search is influenced by individual differences in search proficiency (Boot, Becic, & Kramer, 2009), the adoption of cognitive strategies (Smilek, Dixon, & Merikle, 2006a;Smilek, Enns, Eastwood, & Merikle, 2006b;Watson, Brennan, Kingstone, & Enns, 2010), and environmental factors that make search easy or difficult, including clutter (Smilek, Weiheimer, Kwan, Reynolds, & Kingstone, 2009) and visual eccentricity (Wolfe, 1998). Search also has the advantage of occurring naturally in our everyday lives, whether it involves looking for car keys or locating a book on a shelf.…”

Top-down and bottom-up aspects of active search in a real-world environment.

“…""Splitting"the"search"time"in"this"manner"is"similar"to" the"way"that"the"measurement"of"eye"movements"in"computer?based"visual"search" has"been"used"to"disentangle"different"parts"of"the"reaction"time" "Watson,"et"al.,"2010;"Zelinsky"&"Sheinberg,"1997),"but"here"we" sought"to"describe"the"stages"involved"in"our"unconstrained,"active"search"task."" Figure"3"shows"the"search"time"broken"down"in"this"way.""…”