Purchase this article with an account.
Louis-Alexandre Etezad-Heydari, Hang Zhang, Rob Morris, Laurence Maloney; Planning of saccadic eye movements in an evaluation task. Journal of Vision 2012;12(9):410. doi: https://doi.org/10.1167/12.9.410.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
We investigated eye movements in a simple evaluation task. On each trial, two compositions were sequentially presented for 3s each and observers indicated which of the two was more attractive. Inspired by the painter Kandinsky, we used stochastically-generated compositions that consisted of three large black disks and five small black disks on a white background.
Methods. We tracked observers’ eye movements while they viewed each composition with an Eyelink1000 eyetracker. We systematically varied the centers of gravity (COG) of compositions to see how COG affected the saccadic pattern. We also presented each composition and its three reflections around horizontal and vertical axes. This symmetry manipulation allowed us to assess the relative importance of deterministic and random factors in planning saccades. If saccadic patterns evoked by symmetric compositions were themselves symmetric, we would conclude that saccadic planning is determined by the composition. We characterized the subject’s exploration by computing the convex hull area (CHA) of saccadic endpoints with increasing number of saccades within a composition. Each observer completed 192 = 4x48 trials. Five observers participated.
Results: The mean number of fixations per composition was 8.1.
COG. Surprisingly, the mean location of first saccades did not fall near the COG. Rather it fell on a line joining the center to the COG, approximately 40% percent of the way to the COG, depending on observer. The mean second saccade fell approximately on the COG.
Symmetry. By several metrics, saccadic patterns for the third and later saccades for compositions related by symmetry were markedly more similar than those for unrelated pairs
CH area. As few as four saccades could achieve the maximum possible CHA. Surprisingly, we found a gradual, nearly linear increase of CHA over all saccades. We conjecture that the visual system thereby controls the flow of information into working memory.
Meeting abstract presented at VSS 2012
This PDF is available to Subscribers Only