Search

Pigeons searched for two targets that varied in similarity to items in the background. The targets, simulating 'prey,' were small alphabetic characters presented on computer monitors among distractor characters. In the first experiment, the probability of reinforcement for pecking at the targets was manipulated. In the second experiment, the response requirement for one of the targets was varied. Changing the probability of reinforcement led to more efficient search and increased choice for the more reinforced item. This effect carried over to subsequent equal reinforcement baseline conditions. Increasing the response requirement for one item increased response tune for that item and reduced choice of the item. This effect did not carry over to a subsequent baseline condition. The results suggested that a high probability of reinforcement resulted in improved detection of the item, perhaps through perceptual learning, and increased the incentive that motivated response to the item, but that high response requirement resulted only in an incentive shift.

What is the unit of selection for feature integration in visual search: location or perceptual object? Feature integration theory (A. Treisman, 1988) asserts that it is location. Two alternative models are put forward and tested in a series of 4 experiments using a special conjunctive-search task. In this task, each stimulus item consists of 2 overlapping forms (perceptual objects). In general, the search was more efficient when the search features were linked to the same perceptual object than when they were linked to different perceptual objects at the same stimulus location. This same-object advantage, however, was shown to depend on stimulus discriminability and density, grouping strength, and hierarchical object structure. The results support a hierarchical object-based model, with important implications for feature integration, visual search, late versus early selection, and object-based versus space-based views of attention.

Inhibition-of-return is thought to be a visual search phenomenon characterized by delayed responses to targets presented at recently cued or recently fixated locations. We studied this inhibition effect following the simultaneous presentation of multiple location cues. The results indicated that response inhibition can be associated with as many as four locations at the same time. This suggests that a purely oculomotor account of inhibition-of-return is oversimplified. In short, although oculomotor processes appear to play a role in inhibition-of-return they may not tell the whole story about how it occurs because we can only program and execute eye movements to one location at a time.

It is not always easy to locate routing signs in the visually noisy environment of public buildings. Guidelines to ensure the conspicuity of an object in a complex field are not readily available. In two previous experiments, aiming to provide such guidelines for use in practice, the stimuli were simple simulations of scenes in public buildings, each containing a uniquely colored target (a routing sign) and a varying number of distractors (advertising signs). The results of these two experiments were opposite: in one experiment pop-out of the targets appeared as predicted by current theories of visual search, but in the other experiment performance significantly decreased when the number of distractors increased. Two new experiments are reported here to clarify this discrepancy. From the results it is concluded that pop-out of easy visual targets does not occur on the first few trials of a task. Subjects must have had sufficient practice for pop-out to develop. This finding implies that, contrary to what is common in experiments on visual search, short and heterogeneous series of stimuli should be used to ensure the practical relevance of the results for applications such as sign posting.

Sequential priming refers to speeded visual search when target identity or location is repeated within a trial sequence. In two experiments with pigeons, we addressed the relative contributions of stimulus-driven factors and learned expectancies to this effect. Pigeons pecked at targets during trialwise presentations of visual-search displays. Random-sequence conditions minimized the role of expectancy by introducing same-target or same-location trial sequences unpredictably. Blocked-sequence conditions added predictability by regular repetition of target and/or location over trials. Intertrial interval varied from 0.5 to 3 sec. The findings revealed significant reductions in reaction time during predictable target or location sequences compared with unpredictable repetitions within random contexts. Stimulus-driven factors do not seem to have an important role in many instances of sequential priming. Expectancy-based priming of target and location followed similar patterns.

Models of visual search that assume spatially serial processing of conjunction stimuli break down when displays contain 2 identical conjunction targets: (a) They fail to account for the slopes of the target-absent reaction time (RT)/display size functions, both when the target-present response requires the detection of only 1 of the 2 targets and when it requires the detection of both targets and (b) under redundant target conditions, the probability of 'fast' present RTs is greater than that expected from the independent (serial) processing of the target stimuli. This article shows that G. W. Humphreys and H. J. Muller's (1993) parallel search via recursive rejection (SERR) model can account for human search performance under dual-target conditions. In SERR, simple form conjunctions are coded and grouped (on the basis of their similarity) in a spatially parallel manner, and targets are detected through the activation of a temporary 'template' or through the recursive rejection of grouped distractors.

We consider several ways in which the interpretation of reaction time (RT) data might confound differences in visual search rates with non-search-related factors. To determine whether estimates of search rates for groups differing in age suffered this problem, we compared estimates provided by the RT method with those obtained using a forced-choice method with limited-duration stimuli. The forced-choice method provided faster estimates of search rates. The effects of age, the variable in which we were interested, were comparable, but the difference between results obtained using the two methods suggests the need for caution in using the RT method. We discuss how the forced-choice method can be used, under appropriate circumstances, to provide an independent test of whether subjects are carrying out serial searches and, if they are, to provide search rate estimates even from data obtained using only a single array size.

Asymmetry is demonstrated between conjunctive targets in visual search with no detectable asymmetries between the individual features that compose these targets. Experiment 1 demonstrates this phenomenon for targets composed of color and shape. Experiment 2 and 4 demonstrate this asymmetry for targets composed of size and orientation and for targets composed of contrast level and orientation, respectively. Experiment 3 demonstrates that search rate of individual features cannot predict search rate for conjunctive targets. These results demonstrate the need for 2 levels of representations: one of features and one of conjunction of features. A model related to the modified feature integration theory is proposed to account for these results. The proposed model and other models of visual search are discussed.

The SEarch via Recursive Rejection (SERR) model of visual search is a connectionistmodel that assumes that visual processing of simple form conjunctions proceeds in a spatially parallel manner. It also proposes that search operates through arecursive rejection of these conjunctions. Experiments revealed that the numberof distractors comprising a single group failed to alter the model's performance. When the number of distracting groups was increased, however, the probability of missing targets became greater. The results were found to be identical with functions generated when human subjects searched displays with variable ditractor group numbers.

Young and old Ss were tested in 3 experiments conducted to explore factors leading to age-related performance differences in consistent mapping (CM) and varied mapping (VM) search tasks. The separate and combined influences of memory scanning and visual search on age-related search effects were examined. In both CM letter and CM semantic category search, age interacted with comparison load in visual and hybrid memory--visual search conditions, whereas differential age effects were not present in pure memory search. For VM search, age effects were present only in pure memory search. These data support the view that the separation of type of search training (CM and VM) as well as memory and visual search components is critical for predicting age-related performance differences. The dissociation of the pattern of age effects in memory and visual search suggests that memory and visual search involve different processing mechanisms.

By using a visual search task, this study examined the encoding of orientation and size for stimuli defined in five different surface media: luminance, color, texture, relative motion, and binocular disparity. Results indicated a spatially parallel analysis of size and orientation features for all surface media, with the possible exception of binocular disparity. The data also revealed a search rate asymmetry in the orientation task for all media: Parallel or shallow search functions were obtained for oblique targets in vertical distractors, whereas steeper serial search functions were obtained for vertical targets in oblique distractors. No consistent asymmetry was found for the large and small targets in the size task. There seemed to be common principles of coding in all these different media, suggesting either a single analysis of shape features applied to a common representation or multiple analyses, one for each surface medium, with each extracting a similar set of features. The shared coding principles may facilitate the use of redundancy across media to reduce ambiquities in the locations and shapes of countours in the visual scene.

Search for conjunctions of highly discriminable features can be rapid or even parallel. This article explores three possible accounts based on (a) perceptual segregation, (b) conjunction detectors, and (c) inhibition controlled separately by two or more distractor features. Search rates for conjunctions of color, size, orientation, and direction of motion correlated closely with an independent measure of perceptual segregation. However, they appeared unrelated to the physiology of single-unit responses. Each dimension contributed additively to conjunction search rates, suggesting that each was checked independently of the others. Unknown targets appear to be found only by serial search for each in turn. Searching through 4 sets of distractors was slower than searching through 2. The results suggest a modification of feature integration theory, in which attention is controlled not only by a unitary 'window' but also by a form of feature-based inhibition.

Treisman's Feature Integration Theory and Julesz's Texton Theory explain many aspects of visual search. However, these theories require that parallel processing mechanisms not be used in many visual searchers for which they would be useful, and they imply that visual processing should be much slower than it is. Most importantly, they cannot account for recent data showing that some subjects can perform some conjunction searches very efficiently. Feature Integration Theory can be modified so that it accounts for these data and helps to answer these questions. In this new theory, which we call Guided Search, the parallel stage guides the serial stage as it chooses display elements to process. A computer simulation of Guided Search produces the same general patterns as human subjects in a number of different types of visual search.

During an individual's normal interaction with the environment and other humans, visual and linguistic signals often coincide and can be integrated very quickly. This has been clearly demonstrated in recent eyetracking studies showing that visual perception constrains on-line comprehension of spoken language. In a modified visual search task, we found the inverse, that real-time language comprehension can also constrain visual perception. In standard visual search tasks, the number of distractors in the display strongly affects search time for a target defined by a conjunction of features, but not for a target defined by a single feature. However, we found that when a conjunction target was identified by a spoken instruction presented concurrently with the visual display, the incremental processing of spoken language allowed the search process to proceed in a manner considerably less affected by the number of distractors. These results suggest that perceptual systems specialized for language and for vision interact more fluidly than previously thought.

'Temporal migration' describes a situation in which subjects viewing rapidly presented stimuli (e.g., 9-20 items/s) confidently report a target element as having been presented in the same display as a previous or following stimulus in the sequence. Four experiments tested a short-term buffer model of this phenomenon. Experiments 1 and 4 tested the hypothesis that subjects' errors are due to the demands of the verbal report procedure rather than to perceptual integration. In Experiment 1, 12 color objects were presented at a rate of 9/s. Prior to each sequence, an object was named and subjects responded 'yes' or 'no' to indicate whether the target element (a black frame) occurred with that object. Consistent with the perceptual hypothesis, the yes/no procedure yielded the same results as the verbal report procedure. Experiment 2 tested the hypothesis that the direction of migration depends on 'frame' detection time. Results showed that reaction time to frame detection was significantly faster in trials in which subjects reported the frame on a preceding rather than a following picture. Experiments 3 and 4 used the standard naming procedure and the yes/no procedure to test temporal migration using more complex, interrelated stimuli (objects and scenes). Implications for the use of the temporal migration effect to study visual integration within eye fixations are discussed.

Subjects searched sets of items for targets defined by conjunctions of color and form, color and orientation, or color and size. Set size was varied and reaction times (RT) were measured. For many unpracticed subjects, the slopes of the resulting RT X Set Size functions are too shallow to be consistent with Treisman's feature integration model, which proposes serial, self-terminating search for conjunctions. Searches for triple conjunctions (Color X Size X Form) are easier than searches for standard conjunctions and can be independent of set size. A guided search model similar to Hoffman's (1979) two-stage model can account for these data. In the model, parallel processes use information about simple features to guide attention in the search for conjunctions. Triple conjunctions are found more efficiently than standard conjunctions because three parallel processes can guide attention more effectively than two.

We report evidence for spatially parallel visual search for targets defined by combinations of form elements in visual search. In Section 1, we show that flat search functions occur for combined-form targets when distractor forms are homogeneous and can be grouped together, thus segmenting the target from the distractors. Introducing heterogeneous distractors lessens distractor grouping and can produce serial search. These results cannot be easily attributed to subjects' use of local feature information to discriminate targets. Instead, they suggest that grouping can operate at a level at which combined form information is represented. In Section 2 we show that these grouping effects are spatially scaled by the size of the stimuli. In Section 3 we show that heterogeneity does not prevent flat search functions when the target has a unique defining feature. The data are interpreted in terms of a hierarchial processing system involving both devoted single-feature and combined-feature (junction) maps. Grouping processes can operate at both the single-feature and the combined-form levels. Selection in visual search remains confined to one object description at a time, but this description can be at various spatial scales, including that at the level of grouped forms.