Abstract
Visual working memory is a limited-capacity system which can represent information over brief time periods in the absence of the visual stimuli. How the mechanisms responsible for this working memory maintenance relate to the systems of visual attention that are involved in eye movement generation has long been investigated. However, the vast majority of this research has measured memory performance under a steady-fixation condition compared to other conditions that either required eye movements or intentionally induced eye movements by presenting stimuli. In the current experiment, our goal was to examine the natural occurrence of eye movements in a VWM task in which participants were given no specific instructions regarding eye movement behavior. Participants completed a typical visual working memory task in which a number of colored squares (2, 4, or 6) appeared briefly. After a 1-second delay period, a single item reappeared and participants had to judge whether its color was the same or different as the object previously at that location. In a critical analysis, we examined the mean number of eye movements that occurred during the blank delay period as a function of subsequent accuracy. Trials that were ultimately correct had significantly fewer unprovoked eye movements during the delay than those trials that were ultimately inaccurate. Generally, this finding supports overlap between some component of eye movement generation and/or production and memory maintenance. Such differences in eye movement behavior across trials and between participants may be an important contributor to working memory performance overall. However, the degree to which this behavior is strategic will require further investigation.