Abstract
There are two potential causes of missed targets in visual search tasks: search errors, where the target is never inspected before making a response, and recognition errors, where the target is fixated, but goes unrecognized. Search errors increase with low quitting thresholds; too few items are inspected to reliably find the target. Recognition errors are due to a poor decision process that evaluates whether each scrutinized item is a target or not (and could be due to a conservative criterion, poor evidence accumulation, or a poor search template), resulting in a greater chance of the item failing to reach the target threshold (Wolfe & Van Wert, 2010). Though high working memory has been found to predict greater accuracy in visual search tasks (Schwark, Sandry, & Dolgov, 2013), it is unknown whether the source of this improvement is higher quitting thresholds, better item-by-item decisions, or both. To investigate this issue, eye movements were tracked during a visual search task for a target T among Ls. Following eye-tracking, working memory capacity (WMC) was measured using a change detection task (Luck and Vogel, 1997). We correlated WMC with accuracy, number of items inspected before terminating a trial, and target recognition rate in the visual search task. Results show that working memory is positively correlated with hit rate, replicating previous research. We extend these results by showing higher working memory predicts both an increase in the number of items inspected per trial and a decrease in recognition errors, indicating higher quitting thresholds and better item-by-item decision making. These results could potentially be used to inform screening of new employees whose jobs will be to detect targets in a visual search, such as in the Transportation Security Administration.
Meeting abstract presented at VSS 2016