When visual search functions look like item recognition functions

Visual search data were collected from six Ss on three target set sizes on each of 30 days. Error level was low, and items assigned to memory sets were nonnested and changed from session tosession. For each S. the same item sometimes required a positive and sometimes a negative response (response inconsistency). Combining data over Ss and over successive 6-day blocks, visual search rates as a function of target set size were found to be linear for each of the five 6-day blocks. The slopes of the above functions (memory search time) did not differ significantly over the final four 6-day blocks, and averaged approximately.500 sec per six-character item. These results are qualitatively very similar to results obtained from item recognition studies when error level, memory set structure, degree of response’_ consistency, and practice are handled in the same way in that task. The significantly lower slope obtained on the first 6-day block is shown to be consistent with a speed-accuracy trade off interpretation when error rate is expressed per unit of processing time (percent errors/set size). Over the final three 6-day blocks, where all important parameters ofthe data were highly stable, the intercepts of the memory search functions were found to closely approximate zero, averaging .0068 sec. From this finding, along with the finding that the memory search functions are linear, it is inferred that visual search time is determined entirely by memory search time, or by memory search time and other processes which increase linearly with set size, under the conditions of this experiment. The estimate of memory search time (approximately 83 msec/character) obtained using this visual search procedure is much slower than that obtained using the item recognition procedure (approximately 35\2-40 msec/character). An explanation for this difference is proposed.     

Detection and identification of words and letters in simulated visual search of word lists

Ss were shown a rapid sequence of words and had to (a) make a speeded response to the presentation of a predefined target and (b) report a "response word" which they thought immediately followed the target in the sequence. In Experiment I presentation rate was varied orthogonally with the number of alternative targets. Detection errors and latency increased with target set size, as did the distance of the response word from the target in the list sequence. Increases in presentation rate produced greater target-response distance without affecting detection time. In Experiment II Ss were in some conditions given only the initial letter of the targets; the response could be either the whole word following the target or only its initial letter. The results indicated that Ss could, within limits, concurrently detect initial letters and identify words. Alternatives to hierarchical-type models of stimulus processing in visual search were discussed.     

Searching for many targets: An analysis of speed and accuracy

Three Ss scanned matrices of letters for 40 sessions in a test of Neisser’s claim that feature tests in high-speed searches operate independently and in parallel. In the multiple-target condition (MTC), the matrix contained any one of four target letters, while in the four single-target conditions (STC), the S knew which particular target was embedded in the list. In contrast to previous studies, the error rates for individual target letters in the MTC were analyzed separately rather than being pooled. Two Ss made more errors on the hardest target when searched for in the MTC than in the STC. This difference would be masked by pooling error rates. The third S’s scanning rate in the MTC was not as rapid as in the STC. Neither a sequential nor a strictly parallel feature processing model can account for these data.     

When item recognition and visual search functions are similar

The effect of prolonged practice upon item recognition performance was investigated under conditions of nested positive sets and complete response consistency. Nesting is defined by each positive set containing all the items contained in smaller positive sets. Response consistency is defined by each item in the stimulus set consistently requiring only a positive or only a negative response. A low error level was maintained. Twelve Ss worked with three positive set sizes in each of 36 sessions. Half the Ss worked with digit stimuli and half with pictures. The item recognition function (that function relating response latency and positive set size) was found to be negatively accelerated throughout the course of practice. The effect of positive set size decreased significantly (p < .001) with practice, and set size effects were significantly (p < .03) greater for positive response trials than for negative response trials. Kind of item had no effect on the set size effect. A theoretical framework consistent with these results is suggested. Results from the present study are compared with findings obtained previously from visual search studies. It is concluded that when the procedures in both tasks include response consistency, nested positive sets, and low error levels, the effects of prolonged practice upon the set size from item recognition and visual search are qualitatively very similar.     

Searching for targets: The effects of specific practice

This study contrasts the effects of specific practice vs nonspecific practice in scanning for sets of five-letter targets. Ss were required to search for new targets after extended practice with one specific target set. Simultaneous memory search was not supported, since searching for one item remained faster than for several items. Serial exhaustive memory search was not supported since the search data were nonlinear. There is evidence that the categorization of a target, even after large amounts of specific practice, is not heavily dependent upon its physical features. A strategy change was detected during practice which allowed more efficient search. It is concluded that a single memory search process can account for the data under both high and low levels of specific practice.     

Variables influencing the mode of processing of complex stimuli

The present study attempts to specify some of the conditions under which parallel and serial processing may occur. The three variables studied were (1) type of task, (2) relative set for speed vs accuracy, and (3) practice. Pairs of multidimensional, geometric stimuli were presented either simultaneously or successively to S who was required to indicate whether they were the same or different. Each S participated in nine sessions. For half of the Ss speed was emphasized, and for the other half accuracy was emphasized. The results indicated that: (1) responses were faster with successive presentation than with simultaneous presentation; (2) with successive presentation, processing was serial; (3) in the simultaneous presentation condition, a gradual shift from serial to parallel processing occurred with practice; and (4) the speed and accuracy instructions used in this experiment produced no differential effects on latency or errors.     

Inefficient conjunction search made efficient by concurrent spoken delivery of target identity

Visual search based on a conjunction of two features typically elicits reaction times that increase linearly as a function of the number of distractors, whereas search based on a single feature is essentially unaffected by set size. These and related findings have often been interpreted as evidence of a serial search stage that follows a parallel search stage. However, a wide range of studies has been showing a form of blending of these two processes. For example, when a spoken instruction identifies the conjunction target concurrently with the visual display, the effect of set size is significantly reduced, suggesting that incremental linguistic processing of the first feature adjective and then the second feature adjective may facilitate something approximating a parallel extraction of objects during search for the target. Here, we extend these results to a variety of experimental designs. First, we replicate the result with a mixed-trials design (ruling out potential strategies associated with the blocked design of the original study). Second, in a mixed-trials experiment, the order of adjective types in the spoken query varies randomly across conditions. In a third experiment, we extend the effect to a triple-conjunction search task. A fourth (control) experiment demonstrates that these effects are not due to an efficient odd-one-out search that ignores the linguistic input. This series of experiments, along with attractor-network simulations of the phenomena, provide further evidence toward understanding linguistically mediated influences in real-time visual search processing.     

The role of categorization in visual search for orientation.

Visual search for 1 target orientation is fast and virtually independent of set size if all of the distractors are of a single, different orientation. However, in the presence of distractors of several orientations, search can become inefficient and strongly dependent on set size (Exp. 1). Search can be inefficient even if only 2 distractor orientations are used and even if those orientations are quite remote from the target orientation (e.g. 20 degrees or even 40 degrees away, Exp. 2). Search for 1 orientation among heterogeneous distractor orientations becomes more efficient if the target orientation is the only item possessing a categorical attribute such as steep, shallow (Exp. 3), tilted left or tilted right (Exp. 4), or simply tilted (Exps. 5 and 6). Orientation categories appear to be 1 of several strategies used in visual search for orientation. These serve as a compromise between the limits on parallel visual processing and the demands of a complex visual world.     

Fixed-target efficient search has logarithmic efficiency with and without eye movements

Stage 1 processing in visual search (e.g., efficient search) has long been thought to be unaffected by factors such as set size or lure–distractor similarity (or at least to be only minimally affected). Recent research from Buetti, Cronin, Madison, Wang, and Lleras (Journal of Experimental Psychology: General, 145, 672–707, 2016) showed that in efficient visual search with a fixed target, reaction times increase logarithmically as a function of set size and, further, that the slope of these logarithmic functions is modulated by target–distractor similarity. This has led to the proposal that the cognitive architecture of Stage 1 processing is parallel, of unlimited capacity, and exhaustive in nature. Such an architecture produces reaction time functions that increase logarithmically with set size (as opposed to being unaffected by it). However, in the previous studies, eye movements were not monitored. It is thus possible that the logarithmicity of the reaction time functions emerged simply as an artifact of eye movements rather than as a reflection of the underlying cognitive architecture. Here we ruled out the possibility that eye movements resulted in the observed logarithmic functions, by asking participants to keep their eyes at fixation while completing fixed-target efficient visual search tasks. The logarithmic RT functions still emerged even when participants were not allowed to make eye movements, thus providing further support for our proposal. Additionally, we found that search efficiency is slightly improved when eye movements are restricted and lure–target similarity is relatively high.     

Two-stage model of visual backward masking: Sensory transmission and accrual of effective information as a function of target intensity and similarity

A functional model is used to describe the effect of target intensity and target-set similarity on backward visual masking. The model consists of two distinct stages of visual information processing. The first stage is related to sensory transduction and transmission and is assumed to require a finite and measurable amount of time during which performance-remains at chance. The second stage, associated with central processing, is characterized by a negatively accelerated growth function reflecting the accrual of effective information. Results show that the duration of the transmission stage is inversely related to target intensity. Surprisingly, the rate of information accrual is an interactive function of both target intensity and target-set similarity. The pattern of results is consistent with the interpretation that both intensity and similarity mediate their effect through a common mechanism—the accrual of effective information.     

Electrophysiological evidence for parallel and serial processing during visual search

Event-related potentials were recorded from young adults during a visual search task in order to evaluate parallel and serial models of visual processing in the context of Treisman's feature integration theory. Parallel and serial search strategies were produced by the use of feature-present and feature-absent targets, respectively. In the feature-absent condition, the slopes of the functions relating reaction time and latency of the P3 component to set size were essentially identical, indicating that the longer reaction times observed for larger set sizes can be accounted for solely by changes in stimulus identification and classification time, rather than changes in post-perceptual processing stages. In addition, the amplitude of the P3 wave on target-present trials in this condition increased with set size and was greater when the preceding trial contained a target, whereas P3 activity was minimal on target-absent trials. These effects are consistent with the serial self-terminating search model and appear to contradict parallel processing accounts of attention-demanding visual search performance, at least for a subset of search paradigms. Differences in ERP scalp distributions further suggested that different physiological processes are utilized for the detection of feature presence and absence.     

The effect of a visual indicator on rate of visual search: Evidence for processing control

Search rates were estirnated from response latencies in a visual search task of the type used by Atkinson, Holmgren, and Juola (1969), in which a S searches a small set of letters to determine the presence or absence of a predesignated target. Half of the visual displays contained a marker above one of the letters. The marked letter was the only one that had to be checked to determine whether or not the display contained the target. The presence of a marker in a display significantly increased the estimated rate of search, but the data clearly indicated that Ss did not restrict processing to the marked item. Letters in the vicinity of the marker were also processed. These results were interpreted as showing that Ss are able to exercise some degree of control over the search process in this type of task.     

Attention in visual search: multiple search classes.

Data from visual-search tasks are typically interpreted to mean that searching for targets defined by feature differences does not require attention and thus can be performed in parallel, whereas searching for other targets requires serial allocation of attention. The question addressed here was whether a parallel-serial dichotomy would be obtained if data were collected using a variety of targets representing each of several kinds of defining features. Data analyses included several computations in addition to search rate: (1) target-absent to target-present slope ratios; (2) two separate data transformations to control for errors; (3) minimum reaction time; and (4) slopes of standard deviation as a function of set size. Some targets showed strongly parallel or strongly serial search, but there was evidence for several intermediate search classes. Sometimes, for a given target-distractor pair, the results depended strongly on which character was the target and which was the distractor. Implications from theories of visual search are discussed.     

Attention in visual search: Multiple search classes

Data from visual-search tasks are typically interpreted to mean that searching for targets defined by feature differences does not require attention and thus can be performed in parallel, whereas searching for other targets requires serial allocation of attention. The question addressed here was whether a parallel-serial dichotomy would be obtained if data were collected using a variety of targets representing each of several kinds of defining features. Data analyses included several computations in addition to search rate: (1) target-absent to target-present slope ratios; (2) two separate data transformations to control for errors; (3) minimum reaction time; and (4) slopes of standard deviation as a function of set size. Some targets showed strongly parallel or strongly serial search, but there was evidence for several intermediate search classes. Sometimes, for a given target-distractor pair, the results depended strongly on which character was the target and which was the distractor. Implications from theories of visual search are discussed.     

Categorization and identification of simultaneous targets.

Early and late selection theories of visual attention disagree about whether identification occurs before or after selection. Studies showing the category effect, i.e., the time to detect a letter is hardly affected by the number of digits present in the display, are taken as evidence for late selection theories since these studies suggest parallel identification of all items in the display. As an extension of previous studies, in the present study two categorically different targets were presented simultaneously among a variable number of nontargets. Subjects were shown brief displays of two target letters among either 2, 4 or 6 nontarget digits. Subjects responded 'same' when the two letters were identical and 'different' otherwise. Since the 'same-different' response reflects the combined outcome of the simultaneous targets, late-selection theory predicts that the time to match the target letters is independent of the number of nontarget digits. Alternatively, early-selection theory predicts a linear increase of reaction time with display size since the presence of more than one target disrupts parallel preattentive processing, leading to a serial search through all items in the display. The results provide evidence for the early-selection view since reaction time increased linearly with the number of categorically different nontargets. A control experiment revealed that none of the alternative explanations could account for the display size effect.     

Two-stage model of visual backward masking: sensory transmission and accrual of effective information as a function of target intensity and similarity.

A functional model is used to describe the effect of target intensity and target-set similarity on backward visual masking. The model consists of two distinct stages of visual information processing. The first stage is related to sensory transduction and transmission and is assumed to require a finite and measurable amount of time during which performance remains at chance. The second stage, associated with central processing, is characterized by a negatively accelerated growth function reflecting the accrual of effective information. Results show that the duration of the transmission stage is inversely related to target intensity. Surprisingly, the rate of information accrual is an interactive function of both target intensity and target-set similarity. The pattern of results is consistent with the interpretation that both intensity and similarity mediate their effect through a common mechanism--the accrual of effective information.     

Improvement, learning and retention of skill at visual search

Three experiments examined factors responsible for improvement in visual search. In Experiment I three groups of subjects were each trained for 3000 trials to search for a particular set of target letters among a particular set of background letters. After intervals of 2, 4, or 6 weeks without further practice they were re-tested, either with the same displays or on new displays which they searched for the same target items among new background items. Negative transfer suggested that memory for specific cue-systems distinguishing target from background letters is retained for as long as 4 weeks.

Experiment II examined performance of three very highly practised subjects. After 25 days practice, variations in the size of the target set no longer affected search time. This could not be explained by learning of specific cue systems, since after this amount of practice subjects showed perfect transfer to displays which they searched for the same target items among new background items. Experiment III suggested that practised subjects achieve this high level of competence by learning to make two independent successive decisions, first to locate any member of the target set on the display and next to identify which particular member of the target set has been located.     

“Transient structures”: The effects of practice and distractor grouping on within-dimension conjunction searches

It has been stated that whereas between-dimension (color × orientation) conjunctions can be searched in a “parallel” fashion, within-dimension (color × color) conjunctions are necessarily searched in a “serial self-terminating” fashion (Wolfe et al., 1990). We explored the effects of practice (within 1-h experimental session) and distractor grouping on within-dimension conjunction search tasks. In Experiments 1 and 3, the stimuli were rectangles formed by two adjacent squares; in Experiment 2, the stimuli were plus signs formed by two segments and an intersection. In Experiments 1 and 2, observers were assigned to one of two experimental conditions: In one, all the distractors shared a simple feature (the color blue); in the other, distractors did not share that simple feature. In the first condition, search became more efficient with practice and observers’ performance was characterized by a shallow reaction time (RT) × set size slope; in the second condition, observers’ performance did not improve as much with practice. We propose that the differential effects of practice between these two experimental conditions can be explained in terms of distractor grouping induced by the shared color of the distractors. Experiment 3 showed that, with practice, a shallow RT × set size slope characterized observers’ search for a color × color target among four different distractors that shared a common color. The present results contradict a main tenet of some current visual search models— namely, that within-dimension conjunctions cannot be searched in parallel, and question the validity of using RT × set size slope functions to distinguish between preattentive versus attentive search.     

Guided search: an alternative to the feature integration model for visual search

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.     

Parieto-occipital areas involved in efficient filtering in search: A time course analysis of visual marking using behavioural and functional imaging procedures

Search for a colour-form conjunction target can be facilitated by presenting one set of distractors prior to the second set of distractors and the target: the preview benefit (Watson & Humphreys, 1997). The early presentation of one set of distractors enables them to be efficiently filtered from search. We report two studies investigating the time course of the preview benefit. In Experiment 1 we use a standard reaction time analysis to show that the benefit has a relatively slow time course; old items need to precede the new set by 600 ms or more in order to be fully filtered from search. Furthermore, the reductions in reaction time across time in the preview condition varied nonlinearly with the display size, suggesting that old items were discounted from search in parallel. In Experiment 2 we examined the neural locus of this filtering effect over time, using positron emission tomography (PET). We show that regions of parieto-occipital cortex are selectively activated in a preview search condition relative to a detection baseline. These regions also increase in activation as the preview interval increases (and search then becomes easier), consistent with them modulating the parallel filtering of distractors from targets in spatial search. Interestingly, the same areas as those activated in preview search were also active in conjunction search relative to its own detection baseline. Thus these regions either modulate parallel filtering in conjunction search too, or they modulate different behavioural functions according to task constraints.