Abstract
Purpose In models of visual discrimination, sensory evidence accumulates over time, producing a tradeoff between speed and accuracy. To test whether pursuit and saccades obey the same speed-accuracy tradeoff, we measured the accuracy of pursuit and saccade choices over a large range of latencies. Methods Human observers (n=2) initially fixated a central fixation cross. After a random interval, noise strips (0.7° ver x 40° hor) moving horizontally (14°/s) were presented both above and below (± 2°) fixation. To elicit a range of latencies, the offset of the fixation cross varied in time (+200, 0, −200 ms) relative to the onset of the noise strips. The luminances of the pixels in the two noise strips were drawn from two normal distributions with different means, but with the same standard deviation. The difference in means was adjusted to produce signal strengths of d′ = 0.05 or 0.1. The observers were asked to make an eye movement to and follow the brighter of the two strips. Because the strips moved horizontally in opposed directions and were vertically offset, subjects made a combination of pursuit and saccades on each trial. For each of the two signal strengths, we constructed speed-accuracy curves for pursuit and saccades by plotting cumulative sensitivity as a function of time. Additionally, we measured the sensitivity of pursuit in a 1000-ms perisaccadic interval. Results The trajectories of the speed-accuracy curves for pursuit and saccades were similar — both increased from chance to asymptotic performance by 500 ms. Pursuit reached 95% of its final sensitivity at 44 and 98 ms before the saccade for our two subjects, respectively. Conclusions The similarity in the speed-accuracy tradeoffs for pursuit and saccades supports the idea that choices by both eye movement systems are based on a shared pool of sensory evidence. The perisaccadic enhancement of pursuit sensitivity indicates that this sensory evidence accrues on a similar time scale for both movements.
Supported by NIH EY12212 & The McKnight Foundation (RJK), NSF GRFP (DL), Merck & Cota-Robles (CDC)