|
Visual Search for Flicker
(abstract of the dissertation)
Robert F. Dougherty
University of California, Santa Cruz, CA
Some objects in our visual environment seem to draw our attention to themselves. This stimulus-driven attentional capture involves a combination of salient stimulus properties and the observer's goals and expectations. This study explored attentional capture by flickering stimuli.
Observers were asked to indicate the presence of a target flickering at a different rate than the distractors (experiments 1 and 2) and a target defined by orientation, but with 0-4 of the distractors flickering (experiment 3). The search task was performed using two-interval, forced choice with 8 low spatial frequency, twice detection threshold contrast Gabors arranged concentric at 3-5° from fixation. The Gabors were presented for a variable duration (adjusted with a staircase procedure) and followed by a brief dynamic noise mask. A psychometric function was fit to determine the critical duration necessary for 80% correct. Relevant set size was manipulated in the first two experiments by spatially precueing 2, 4 or 8 possible target locations.
The first experiment showed that low target-distractor discriminability generally produced longer critical durations that increased as relevant set size increased. For more discriminable targets, search times were independent of set size. The second experiment showed that even with equal target-distractor discriminability, search for rapid flicker was faster than search for slower flicker. While high frequency targets can be found faster than low frequency targets when observers are explicitly looking for flicker, the third experiment demonstrated that the presence of 1-4 flickering distractors does not hinder performance in a search-for-orientation task.
This study follows previous work in demonstrating the importance of stimulus discriminability in visual search tasks. Only when discriminability fails to explain search data must other mechanisms be postulated. The failure of discriminability to fully explain these data can be accounted for if we assume that rapid flicker captures attention when observers are explicitly searching for flicker. This interpretation suggests that a rapid flicker channel (postulated by current models of flicker detection and possibly instantiated in the magnocellular system) mediates the attentional capture effects of motion and abrupt stimulus onset.