| |
Abstract:
The antisaccade task requires making saccades in the
direction opposite to a peripherally-flashed cue. We examine the
predictions of a functional-neuroanatomical model that emphasizes
the timing of competing saccade programs for determining whether
one makes a reflexive error or an antisaccade. As the number of
on-line parameters for programming an antisaccade increases,
antisaccade programming slows, which shifts timing dynamics further
in favor of the competing reflexive saccade. Our earlier work
partially supported this model: decreasing the number of parameters
for on-line antisaccade programming improved performance.
Decreasing uncertainty in the number of possible cue locations
(from 2 to 1) lead to near perfect performance. Increasing the
number of possible cue locations (2 to 4), however, did not affect
error rate when cues were positioned along a single axis. In two
studies we examine the hypothesis that on-line programming of
saccadic amplitude does not increase the timing of antisaccade
programming or increase error rate, but that increasing the
possible directions of antisaccades will increase reflexive errors.
In study 1, errors in a 2-choice antisaccade task went from 29%
when the cue direction varied, to 4% when possible targets varied
in amplitude only. In study 2, when saccade direction varied,
errors increased from 27% to 43% in a 2- versus 4- choice
antisaccade task. Changes in amplitude only did not affect error
rate. These results are examined in the context of a competitive
timing model.
|
