Even in correctable search, some types of rare targets are frequently missed

Socially important visual search tasks, such as airport baggage screening and tumor detection, place observers in situations where the targets are rare and the consequences of failed detection are substantial. Recent laboratory studies have demonstrated that low target prevalence yields substantially higher miss errors than do high-prevalence conditions, in which the same targets appear frequently (Wolfe, Horowitz, & Kenner, 2005; Wolfe et al., 2007). Under some circumstances, this \ldprevalence effect\rd can be eliminated simply by allowing observers to correct their last response (Fleck & Mitroff, 2007). However, in three experiments involving search of realistic X-ray luggage images, we found that the prevalence effect is eliminated neither by giving observers the choice to correct a previous response nor by requiring observers to confirm their responses. This prevalence effect, obtained when no trial-by-trial feedback was given, was smaller than the effect obtained when observers searched through the same stimuli but were given trial-by-trial feedback about accuracy. We suggest that low prevalence puts pressure on observers in any search task, and that the diverse symptoms of that pressure manifest themselves differently in different situations. In some relatively simple search tasks, misses may result from motor or response errors. In other, more complex tasks, shifts in decision criteria appear to be an important contributor.     

Rare targets are less susceptible to attention capture once detection has begun

Rare or low probability targets are detected more slowly and/ or less accurately than higher probability counterparts. Various proposals have implicated perceptual and response-based processes in this deficit. Recent evidence, however, suggests that it is attentional in nature, with low probability targets requiring more attentional resources than high probability ones to detect. This difference in attentional requirements, in turn, suggests the possibility that low and high probability targets may have different susceptibilities to attention capture, which is also known to be resource-dependent. Supporting this hypothesis, we found that, once attentional resources have begun to be engaged by detection processes, low, but not high, probability targets have a reduced susceptibility to capture. Our findings speak to several issues. First, they indicate that the likelihood of attention capture occurring when a given task-relevant stimulus is being processed is dependent, to some extent, on how said stimulus is represented within mental task sets. Second, they provide added support for the idea that the behavioural deficit associated with low probability targets is attention-based. Finally, the current data point to reduced top-down biasing of target templates as a likely mechanism underlying the attentional locus of the deficit in question.     

Visual search asymmetry with uncertain targets

The underlying mechanism of search asymmetry is still unknown. Many computational models postulate top-down selection of target-defining features as a crucial factor. This feature selection account implies, and other theories implicitly assume, that predefined target identity is necessary for search asymmetry. The authors tested the validity of the feature selection account using a singleton search task without a predefined target. Participants conducted a target-defined and a singleton search task with a circle (O) and a circle with a vertical bar (Q). Search asymmetry was observed in both tasks with almost identical magnitude. The results were not due to trial-by-trial feature selection, because search asymmetry persisted even when the target was completely unpredictable. Asymmetry in the singleton search was also observed with more complex stimuli, Kanji characters. These results suggest that feature selection is not necessary for search asymmetry, and they impose important constraints on current visual search theories.     

Search for multiple targets: remember the targets, forget the search

Models of visual search performance typically assume that search proceeds by sampling without replacement. This requires memory for each deployment of attention. We tested this assumption of memory-driven search using a multiple-target search paradigm. We held total set size constant, varied the number of targets in the display, and asked subjects to report whether or not there were at least n targets present, where n was varied by block. This allowed us to measure the time to find each subsequent target. Memory-driven search predicts that reaction time should be a linear function of n. The alternative memory-free search hypothesis predicts an accelerating function. The data falsify the memory-driven hypothesis. They were consistent with the memory-free search hypothesis but would also be consistent with memory for a small number of previously attended locations.     

Intra- and cross-dimensional visual search for single-feature targets

Cross-dimensional visual search for single-feature targets that differed from the distractors along two dimensions (color and orientation) was compared with intradimensional search for targets that differed from the distractors along a single dimension (either orientation or color). The design of the first three experiments differed from those of previous studies in that participants were required to respond differently to each of the targets. Experiments 1-3 were similar except that in Experiment 1, the distractors were homogeneous; in Experiment 2, two types of distractors were used in equal proportions; and in Experiment 3, two types of distractors were used but one of the distractors was a singleton. The findings, contrary to those of previous studies, revealed that cross-dimensional search is at least as efficient and for some targets even more efficient than intradimensional search. These results suggest that the details of stimulus-to-response mapping are essential in comparing intra- and cross-dimensional tasks. Experiment 4 used a priming design and did not support an explanation based on grouping processes. We outline an explanation for all the findings based on a recent cross-dimensional response selection model by Cohen and Shoup (1997).     

False feedback increases detection of low-prevalence targets in visual search

Many critical search tasks, such as airport and medical screening, involve searching for targets that are rarely present. These low-prevalence targets are associated with extremely high miss rates Wolfe, Horowitz, & Kenner (Nature, 435, 439?C440, 2005). The inflated miss rates are caused by a criterion shift, likely due to observers attempting to equate the numbers of misses and false alarms. This equalizing strategy results in a neutral criterion at 50?% target prevalence, but leads to a higher proportion of misses for low-prevalence targets. In the present study, we manipulated participants?? perceived number of misses through explicit false feedback. As predicted, the participants in the false-feedback condition committed a higher number of false alarms due to a shifted criterion. Importantly, the participants in this condition were also more successful in detecting targets. These results highlight the importance of perceived prevalence in target search tasks.     

Incidental visual memory for targets and distractors in visual search

We explored incidental retention of visual details of encountered objects during search. Participants searched for conjunction targets in 32 arrays of 12 pictures of real-world objects and then performed a token discrimination task that examined their memory for visual details of the targets and distractors from the search task. The results indicate that even though participants had not been instructed to memorize the objects, the visual details of search targets and distractor objects related to the targets were retained after the search. Distractor objects unrelated to the search target were remembered more poorly. Eye-movement measures indicated that the objects that were remembered were looked at more frequently during search than those that were not remembered. These results provide support that detailed visual information is included incidentally in the visual representation of an object after the object is no longer in view.     

Eye movements during search for coded and uncoded targets

Eye movements and search time of four subjects were studied as they searched for a target dial in a 4 х 4 array of dials which were differentiated by (1) color, (2) shape, (3) a combination of color and shape, or (4) were uncoded. Subjects did not exhibit a characteristic scanpath, and method of scanning did not relate to search time. Search time varied reliably among conditions; it was generally shortest in the color condition, followed by color shape, shape, and the uncoded condition. Subjects were capable of using both shape and color simultaneously. Search time was strongly associated with the average number of fixations required for target detection but not with other measures of eye movements. Fixation duration was a particularly inconsistent measure.     

Found and missed: failing to recognize a search target despite moving it

We present results from five search experiments using a novel 'unpacking' paradigm in which participants use a mouse to sort through random heaps of distractors to locate the target. We report that during this task participants often fail to recognize the target despite moving it, and despite having looked at the item. Additionally, the missed target item appears to have been processed as evidenced by post-error slowing of individual moves within a trial. The rate of this 'unpacking error' was minimally affected by set size and dual task manipulations, but was strongly influenced by perceptual difficulty and perceptual load. We suggest that the error occurs because of a dissociation between perception for action and perception for identification, providing further evidence that these processes may operate relatively independently even in naturalistic contexts, and even in settings like search where they should be expected to act in close coordination.     

Target-tilt and vertical-hemifield asymmetries in free-scan search for 3-D targets

In this study, asymmetries in finding pictorial 3-D targets defined by their tilt and rotation in space were investigated by means of a free-scan search task. In Experiment 1, feature search for cube tilt and rotation, as assessed by a spatial forced-choice task, was slow but still exhibited a characteristic “flat” slope; it was also much faster to upward-tilted cubes and to targets located in the upper half of the search field. Faster search times for cubes and rectangular solids in the upper field, an advantage for upward-tilted cubes, and a strong interaction between target tilt and direction of lighting (upward or downward) for the rectangular solids were all demonstrated in Experiment 2. Finally, an advantage in searching for tilted cubes located in the upper half of the display was shown in Experiment 3, which used a present-absent search task. The results of this study confirm that the upper-field bias in visual search is due mainly to a biased search mechanism and not to the features of the target stimulus or to specific ecological factors.     

The target effect: Visual memory for unnamed search targets

Search targets are typically remembered much better than other objects even when they are viewed for less time. However, targets have two advantages that other objects in search displays do not have: They are identified categorically before the search, and finding them represents the goal of the search task. The current research investigated the contributions of both of these types of information to the long-term visual memory representations of search targets. Participants completed either a predefined search or a unique-object search in which targets were not defined with specific categorical labels before searching. Subsequent memory results indicated that search target memory was better than distractor memory even following ambiguously defined searches and when the distractors were viewed significantly longer. Superior target memory appears to result from a qualitatively different representation from those of distractor objects, indicating that decision processes influence visual memory.     

Target-tilt and vertical-hemifield asymmetries in free-scan search for 3-D targets.

In this study, asymmetries in finding pictorial 3-D targets defined by their tilt and rotation in space were investigated by means of a free-scan search task. In Experiment 1, feature search for cube tilt and rotation, as assessed by a spatial forced-choice task, was slow but still exhibited a characteristic "flat" slope; it was also much faster to upward-tilted cubes and to targets located in the upper half of the search field. Faster search times for cubes and rectangular solids in the upper field, an advantage for upward-tilted cubes, and a strong interaction between target tilt and direction of lighting (upward or downward) for the rectangular solids were all demonstrated in Experiment 2. Finally, an advantage in searching for tilted cubes located in the upper half of the display was shown in Experiment 3, which used a present-absent search task. The results of this study confirm that the upper-field bias in visual search is due mainly to a biased search mechanism and not to the features of the target stimulus or to specific ecological factors.     

On the relation between identifying and locating masked targets in visual search

Three experiments were designed to determine whether naming is contingent on locating in a visual search task. Subjects were required to identify a masked target whose location was known (I|L) or unknown (I) and to locate a masked target whose identity was known (L|I) or unknown (L). The location-contingent hypothesis predicts a relationship among the tasks such that P(L) P(I|L) = P(I), since P(I) and P(L) P(I|L) both estimate the joint probability of identifying and locating the target (i.e. P(IλL)). This relationship held in Experiment I where targets were presented alone, and in Experiment II where targets were presented with dots as noise elements, but not in Experiment III where Xs were noise elements. The results are discussed in terms of the generality of the location-contingent hypothesis.     

Sharp targets are detected better against a figure, and blurred targets are detected better against a background

There is growing evidence that the performance of perceptual tasks is often facilitated by perceived "figureness." Accuracy in detection and discrimination of targets is higher when the targets are presented in figural regions than when they are presented in ground regions of an image. This "figure superiority" might be a result of a functional specialization in the visual analysis of figure; recent theories have also assumed a functional specialization in the visual analysis of ground. If so, we might expect "ground superiority" in situations where task performance requires information available primarily through analysis of ground. We manipulated the spatial frequency of a small line segment and found that when it was sharp (i.e., the high-spatial-frequency components were present), it was detected better in figural regions, but when we blurred it (only the low-to-medium spatial frequencies were present) it was detected better in ground regions. These findings support the view that figure and ground analyses involve different specialized functions.     

Asymmetry in visual search for targets defined by differences in movement speed.

Perception of motion speed was investigated with the visual search paradigm, using human Ss. When searching for a fast target among slow distractors, reaction time was minimally affected as the number of distractors was increased. In contrast, reaction time to detect a slow target among fast distractors was slow and linearly related to the number of distractors. The effect cannot be attributed to differences in temporal frequency, discriminability, or one type of representation that might result from spatiotemporal filtering. An alternative hypothesis that can account for the asymmetry is that speed detectors operate as high-pass filters in the velocity domain. This hypothesis is in agreement with results obtained in psychophysical studies on motion adaptation as well as data from single-cell recordings in nonhuman species.