Purchase this article with an account.
Annie Tran, James Hoffman; Visual Attention is Required for Multiple Object Tracking. Journal of Vision 2013;13(9):1282. doi: 10.1167/13.9.1282.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Pylyshyn and Storm (1988) proposed that observers track multiple objects by assigning pointers to the targets. These pointers are thought to be part of a preattentive mechanism that is independent of attention, and allow observers to reference the location of the targets without providing any information about its properties or identity. Access to object properties is provided by allocating attention to that object. An alternative explanation for multiple object tracking is the "spotlight of attention" model. This model holds that attention is a flexible resource that can be allocated to each target, a process that is necessary for object tracking (Alvarez & Franconeri, 2007). In our experiment, we attempted to determine whether visual attention is required for tracking. Observers were required to track multiple moving objects while trying to discriminate a letter target appearing at a random time on one of the objects. The letter was presented by removing segments of a figure eight object, thus reducing onset cues. In different blocks of trials, we systematically varied the probability of the letter appearing in tracked objects vs. distractors. If attention is not required for tracking, then attention should be preferentially allocated to objects that are likely to contain the search target. However, if attention is necessary for tracking, then good tracking performance should be accompanied by superior detection of letters appearing on tracked objects compared to distractors and this effect should be independent of the probability manipulation. Our results indicated that the probability manipulation had no effect on observers' letter discrimination speed, F(2, 18) = 1.63, p = .23. Letters on tracked objects were consistently identified faster relative to distractors, F(1, 9) = 34.72, p<.001 and this effect did not interact with probability, F(2, 18) = 2.23, p = .16. Allocation of attention to tracked objects appears to be obligatory.
Meeting abstract presented at VSS 2013
This PDF is available to Subscribers Only