Purchase this article with an account.
Richard F. Murray, Allison B. Sekuler, Patrick J. Bennett; A linear cue combination framework for understanding selective attention. Journal of Vision 2003;3(2):2. doi: https://doi.org/10.1167/3.2.2.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Using a linear cue combination framework, we develop a measure of selective attention that describes the relative weight that an observer assigns to attended and unattended parts of a stimulus when making perceptual judgments. We call this measure attentional weight. We present two methods for measuring attentional weight by calculating the trial-by-trial correlation between the strength of attended and unattended parts of a stimulus and the observer’s responses. We illustrate these methods in three experiments that investigate whether observers can direct selective attention according to contrast polarity when judging global direction of motion or global orientation. We find that when observers try to judge the global direction or orientation of the parts of a stimulus with a given contrast polarity (white or black), their responses are nevertheless strongly influenced by parts of the stimulus that have the opposite contrast polarity. Our measure of selective attention indicates that the influence of the opposite-polarity distractors on observers’ responses is typically 65% as strong as the influence of the targets in the motion task, and typically 25% as strong as the targets in the orientation task, demonstrating that observers have only a limited ability to direct attention according to contrast polarity. We discuss some of the advantages of using a linear cue combination framework to study selective attention.
This table shows the mean total rightward displacement of the target and distractor dots, conditional on the target signal dots moving left or right and the observer responding “left” or ”right.” For example, the top left entry shows that for observer A.N.C. in the 50L50L condition, the average total target dot displacement was 2.364 deg to the right on trials where the target signal dots moved right and the observer responded ”right.” The values in this table are the coordinates of the conditional mean displacements shown in Figures 5 and 6 as red and green dots. Note that both the target and the distractor displacements were correlated with observers’ responses: all mean displacements were further to the right when observers responded ”right” than when observers responded “left.” This was true even in the 50L50D condition, where observers tried to ignore the distractor dots.
The first two columns of numbers show the proportion of trials on which the observer responded “right,” conditional on the target and distractor signal dots moving left or right. for example, the top left cell shows that observer A.J.R. responded “right” on 75% of the trials on which both the target and the distractor signal dots moved to the right. The third and fourth columns show the attentional weight k and the target sensitivity d′T calculated from these conditional response probabilities using the methods described in the text. The error values are SEs.
See notes for Table 2.
CW = clockwise; CCW = counterclockwise. See caption of Table 2 for details.
This PDF is available to Subscribers Only