Individual differences in gaze behavior: Comparing high-level and sensory contributions

Poster Presentation 43.409: Monday, May 22, 2023, 8:30 am – 12:30 pm, Pavilion
Session: Eye Movements: Individual differences, novel measurement

Dyllan Simpson1 (), Benjamin Wolfe1, Anna Kosovicheva1; 1University of Toronto Mississauga

Individual differences in gaze behavior are well-established across a range of tasks and stimuli, from examining natural scenes to visual search, but their source is less well-understood. Some differences may arise from variation in high-level strategies for a specific visual task, while others may arise from individual differences in sensory processing (e.g., variation in visual crowding). To examine the contributions of these factors to individual differences in scan paths, we asked participants to perform one of two tasks (a visual search task or an ensemble perception task) on the same stimulus array, consisting of a set of radial frequency patterns arranged in an annulus around an initial central fixation. In the ensemble task, participants reported the average amplitude of the patterns, and in the visual search task, they reported whether a four-pointed shape was present among five-pointed shapes. In each task, we found large individual differences in scan paths between participants, with large variability in initial saccade direction. As expected, in each task, we observed higher trial-to-trial correlations in scan paths within than between participants. These correlations were affected by task alone, since the stimuli were identical; trial-to-trial correlations were higher within participants in our ensemble task than in our search task, indicating that strategies based on task demands contribute to these differences. Finally, to measure the contributions of differences in peripheral vision, we separately measured local differences in the strength of crowding for each participant. We measured critical spacing at a set of 12 peripheral locations around the annulus, comparing these values to participants’ initial saccade direction. This revealed a correlation (r = .21) between the proportion of initial saccades to a given location and critical spacing at the corresponding location. Together, these results suggest that high-level differences and sensory differences contribute differently to idiosyncrasies in participants’ gaze behavior.

Acknowledgements: This work was supported by a UTM Undergraduate Research Grant to D.S. and an NSERC Discovery Grant (RGPIN-2022-03131) to A.K.