Yehoshua Tsal

Perceptual load as a major determinant of the locus of selection in visual attention.

In this paper, we propose that the debate concerning the locus of attentional selection can be resolved by specifying the conditions under which early selection is possible. In the first part, we present a theoretical discussion that integrates aspects from structural and capacity approaches to attention and suggest that perceptual load is a major factor in determining the locus of selection. In the second part, we present a literature review that examines the conditions influencing the processing of irrelevant information. This review supports the conclusion that a clear physical distinction between relevant and irrelevant information is not sufficient to prevent irrelevant processing; early selection also requires that the perceptual load of the task be sufficiently high to exceed the upper limit of available attentional resources.

Location dominance in attending to color and shape.

Four experiments investigated whether Ss direct attention to stimulus location when attempting to attend to its color or shape. In the first 2 experiments a given property (location, color, or shape) of a letter cue instructed Ss whether to report any letters from a subsequent display. Regardless of which property was relevant, Ss reported letters adjacent to the cue and not those similar to its color or shape. In the last 2 experiments, the varied location of a cue was irrelevant to the task, whereas its varied color instructed Ss to report a letter in a given location or of a given shape. Targets adjacent to the cue were reported faster than those remote from the cue. The results suggest that attempting to attend to any aspect of a stimulus entails directing attention to its location.

Attending to color and shape: the special role of location in selective visual processing.

Subjects were presented with circular arrays of letters and were instructed to report first a given target (or targets) and then any other letters they could identify. The targets) was (were) a letter of a given color (Experiment 1) or a given shape (Experiment 2), or two letters of a given shape (Experiment 3). In all three experiments, the additional letters reported tended to be adjacent to the first reported target(s). The results suggest that the selective processing of targets specified by color or by shape is accomplished by attending to the targets’ locations.

Diluting the burden of load: perceptual load effects are simply dilution effects.

The substantial distractor interference obtained for small displays when the target appears alone is reduced in large displays when the target is embedded among neutral letters. This finding has been interpreted as reflecting low-load and high-load processing, respectively, thereby supporting the theory of perceptual load (Lavie & Tsal, 1994). However, a possible alternative interpretation of this effect is that the distractor is similarly processed in both displays, yet its interference in the large ones is diluted by the presence of the neutral letters. We separated the effects of load and dilution by introducing dilution displays. They contained as many letters as the high-load displays but were clearly distinguished from the target, thus allowing for a low-load processing mode. Distractor interference obtained under both the low-load and high-load conditions disappeared under the dilution condition. Hence, the display size effect traditionally misattributed to perceptual load is fully accounted for by dilution. Furthermore, when dilution is controlled for, it is high load not low load producing greater interference.

The Diversity of Attention Deficits in ADHD The Prevalence of Four Cognitive Factors in ADHD Versus Controls

The performance of participants with attention-deficit/hyperactivity disorder (ADHD) relative to control participants was measured on four tasks uniquely assessing the functions of selective attention, executive attention, sustained attention, and orienting of attention. The results showed that deficits in sustained attention were the most pronounced, characterizing most participants with ADHD and deficits in each of the other three functions characterized more than half of these participants. Different participants with ADHD revealed different clusters of attentional deficits. These results call for a revision of leading theories of ADHD that identify the core of the pathology as a sole deficit in executive functions.

Inattention Magnifies Perceived Length: The Attentional Receptive Field Hypothesis

Five experiments demonstrated that a briefly presented vertical line is judged as longer when it is unattended relative to when it is attended. This effect was obtained in estimating the length of 1 of 5 possible lines (Experiment 1) and in matching the length of a test line to a criterion line (Experiments 3 and 4). The directional effect of attention was eliminated when participants estimated the length difference between 2 simultaneously presented lines (Experiment 2). An additional matching experiment (Experiment 5) demonstrated similar lengthening effects for unattended lines and for unattended distances separated by vertically displaced dots. It is proposed that the metric for unattended stimuli is composed of large attentional receptive fields and that the final output is mediated by rounding up processes, so that the unattended line is systematically perceived as longer than the attended one.

Object features, object locations, and object files: Which does selective attention activate and when?

The representation within which attention operates was investigated in 3 experiments. The task was similar to that of R. Egly, J. Driver, and R. D. Rafal (1994). Participants had to detect the presence of a target at 1 of 4 ends of 2 shapes, differing in color and form. A precue appeared at 1 of the 4 possible corners. The 2 shapes occupied either the same or different locations in the cuing and target displays. The results showed that the cued object location was attended whether or not space was task relevant, whereas the cued object features (color and form) were attended only when these were task relevant. Moreover, when object file continuity was maintained through continuous movement, attention was found to follow the cued object file as it moved while also accruing to the cued location.

Do Illusory Conjunctions Support the Feature Integration Theory? A Critical Review of Theory and Findings

Presents a critical review of the feature integration theory and the studies of Treisman and Schmidt ( 1982); Prinzmetal, Presti, and Posner ( 1986); and Briand and Klein (1987) and suggests that the phenomenon of illusory conjunctions does not support Treismans theory of feature integration. I propose that the theory is too vague because it does not explicate the processes that glue features into objects and that each of the reviewed studies has suffered from methodological difficulties that leave the data open to alternative interpretations. The only solid demonstration that attention facilitates feature integration is provided by Experiment 3 of Prinzmetal et al.s study. This finding, however, is irrelevant to the question of whether feature perception and feature integration can or cannot be performed preattentively. It may simply suggest that in addition to its effect on feature perception, attention can also influence feature integration.

Where have we gone wrong? Perceptual load does not affect selective attention

The theory of perceptual load (Lavie & Tsal, 1994) proposes that with low load in relevant processing left over resources spill over to process irrelevant distractors. Interference could only be prevented under High-Load Conditions where relevant processing exhausts attentional resources. The theory is based primarily on the finding that distractor interference obtained in low load displays, when the target appears alone, is eliminated in high load displays when it is embedded among neutral letters. However, a possible alternative interpretation of this effect is that the distractor is similarly processed in both displays, yet its interference in the large displays is diluted by the presence of the neutral letters. We separated the possible effects of load and dilution by adding dilution displays that were high in dilution and low in perceptual load. In the first experiment these displays contained as many letters as the high load displays, but their neutral letters were clearly distinguished from the target, thereby allowing for a low load processing mode. In the second experiment we presented identical multicolor displays in the Dilution and High-Load Conditions. However, in the former the target color was known in advance (thereby preserving a low load processing mode) whereas in the latter it was not. In both experiments distractor interference was completely eliminated under the Dilution Condition. Thus, it is dilution not perceptual load affecting distractor processing.

The role of segmentation in prospective and retrospective time estimation processes.

In five experiments, we investigated the effects of the segmentation level of an interval on its perceived duration. A prospective paradigm and an absolute time estimation method were used in two experiments, and in two others we used a retrospective paradigm and a comparative estimation method. A positive relationship was obtained between segmentation level of the estimated interval and its perceived duration under retrospective-comparative conditions for both auditory and tactual stimuli, but no relationship was found under prospective-absolute conditions. The paradigm, estimation method, and segmentation level were jointly manipulated in the fifth experiment. The impact of segmentation was significant under retrospective (both absolute and comparative) and close to significant under prospective-comparative conditions. These findings suggest that high-priority events are perceived and coded as contextual changes and that the impact of segmentation on time estimation is mediated by memory processes.