The Dynamics of Joint Encoding for Visual Working Memory

Abstract

How do people decide when and how to collaborate during cognitively demanding tasks? This thesis investigates the dynamics of joint visual working memory, focusing on the mechanisms by which individuals coordinate attention and form collaborative strategies under varying task constraints. Utilizing a novel spatial working memory task in which pairs of participants viewed a grid of images, we examined when participants chose to collaborate, whether they adopted spatial specialization strategies, and how these behaviors differed according to changes in task complexity. Participants completed a series of trials in both solo and dyadic conditions, each involving 10 seconds to encode images hidden under tiles across one of five grid sizes (4, 8, 16, 24, or 36 images). Results showed that participants performed better in dyads than alone, most notably within mid-sized grid conditions. Dyads who adopted collaborative strategies – such as dividing the grid into regions – tended to form these conventions early, even without verbal communication. However, not all participants collaborated effectively, as some displayed no significant behavioral differences between solo and dyadic task performance. Together, these results suggest that humans can tacitly arrive at collaborative conventions to split up cognitively demanding tasks, and can moderately calibrate their strategies based on the potential benefits of collaboration on memory performance.