Purchase this article with an account.
D. McConnell, G. Grudic, D. Knill, V. Kumar; Reach corrections to unnoticed target perturbations. Journal of Vision 2001;1(3):254. doi: 10.1167/1.3.254.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: We measured the dynamics of subjects' unconscious corrections in pointing movements made in response to unnoticed changes in target position during saccades. Methods: Observers performed simple pointing movements to a small target projected onto a tabletop. At the beginning of each trial, subjects placed their fingertips at a fixed starting location, located at the near edge of the table. The target for a reach was displayed at one of 3 positions placed along an invisible semi-circular arc 19.5 cm away form the starting point. Subjects fixated a fixation mark projected on the table 30 cm to the left or right of the target location. At the sound of a tone subjects simultaneously shifted their gaze and made a fast pointing movement (∼600 msec.) to the target. An EOG eye-tracking device detected the onset of the saccade. On two-thirds of the trials, when the beginning of a saccade was detected, the target was moved 1cm to the left or right. The hand was visible under normal room illumination. Subjects reported only rarely detecting the perturbation (on ∼1% of trials). An Optotrak system measured the three-dimensional motion of subjects' pointing finger at a sampling rate of 500 Hz. Modeling: Finger velocity was modeled as the sum of an autoregressive process (dependent on the previous movement history) and a time-varying, weighted contribution of the target perturbation. The weight assigned to the target perturbation provides a measure of the time-varying influence of the target perturbation on subjects' movements. Results: Model fits to the data showed a temporally localized influence of the perturbation approximately 200 msecs after the end of a saccade. Discussion: Unconscious corrections in pointing movements to unnoticed target perturbations occur with a reaction time of ∼ 200 msecs, significantly larger than reaction times to detectable perturbations. Moreover, the corrective movement is localized to a small (∼40 msec) window of time.
This PDF is available to Subscribers Only