Visual exploration goes down with higher working memory load, while more saccades negatively impact recall

Poster Presentation: Tuesday, May 21, 2024, 8:30 am – 12:30 pm, Banyan Breezeway
Session: Visual Memory: Working memory and attention

Amit Rawal1,2 (), Rosanne L. Rademaker1; 1Ernst Strüngmann Institute for Neuroscience in Cooperation with the Max Planck Society, Frankfurt, Germany, 2Faculty of Behavioural and Human Movement Sciences, Vrije Universiteit Amsterdam, The Netherlands

During active visual exploration and engagement with our daily environment, we can also entertain many thoughts. Given a limited processing capacity, a balance must be struck between how much we prioritise the information held in mind, and how much we explore the external world. With rising memory demands, might the tendency to sample new inputs go down? Conversely, might visual exploration impact how well a memorised item is recalled? To investigate how internal memory maintenance and external exploration interact, we varied working memory load while participants freely viewed images with varying semantic content. While we tracked their gaze, 20 participants remembered 1, 3, or 5 randomly oriented gratings for 5.5s until cued to report one item via method-of-adjustment. During the delay, participants could freely view an intact natural scene, or a locally phase-scrambled natural scene, for 5s. As expected, recall error rises monotonically with increasing set size. Semantic content also impacted recall, with scrambled images resulting in smaller errors than natural images. To explore if memory load impacts ocular behaviour, we looked at scanpath length (total distance travelled) and number of saccades. Load had no effect on the number of saccades, but higher load did lead to shorter scanpaths. Natural scenes induced more saccades and longer scanpaths compared to scrambled scenes. To explore if ocular behaviour is related to memory performance, we median-split trials by recall error, and found a larger number of saccades on trials with poorer recall, but no association between performance and scanpath length. This was true irrespective of load or semantic content, and complements our previously discovered positive association between recall error and saccade rate during fixation. In sum, we propose a bi-directional link between ocular behaviour and memory recall, where mental contents impact visual exploration, and the amount of ocular activity impacts recall.