Abstract
Working memory resources are incredibly limited yet every day we must maximize the goal-relevant information we encode from our environment. Working memory, however, is not in complete isolation. Prior knowledge in long-term memory can aide in maximizing the information we encode. Here we used a full-report procedure in a visual working memory paradigm to examine the influence of prior information on resource allocation in working memory. For each trial, six colored circles appeared at a random angle about a fixed radius. The location of five of the six colors was drawn from a uniform distribution (“non-prior” items), while the remaining color was drawn from a von Mises distribution with a standard deviation of 20 degrees (“prior” item). The color and mean of the prior distribution were randomly determined for each participant at the beginning of the experiment and remained constant. Participants first completed a training phase where they were explicitly shown and tested on both the color and location of the prior distribution. For the remainder of the experiment, participants reported the locations of all six colors in any order on each trial. To verify participants maintained the prior information, they were tested on the color and location of the prior item at the end of each run. We found that participants allocated fewer resources to the prior items, as evidenced by 1) a bias to report the prior item later in the response sequence (typically 4th of 6 positions), and 2) a decrease in precision for reports of non-prior items that appeared near the prior location. Together these findings show that participants are using strategies to prioritize encoding items with no prior information. These results give us a better understanding of how working memory may rely on long-term memory to strategically encode information from our environment.