Abstract
Previous studies have shown that attention shifts to the saccade goal before movement onset (e.g. Deubel et al., 2008), with a peak approximately 50 ms before saccade onset. The current experiment investigated the timing and dynamics of this attention shift during anti-saccade planning; we investigated whether attention shifts only to the anti-saccade goal location (SG) or to the target location (TL) or both, and when. To do so, we asked participants to perform a dual-task paradigm (similar to the one used by Deubel & Schneider, 1996), involving performing blocks of either anti-saccades or pro-saccades as well as discriminating symbols (DS) at the same time. The DS appeared during the saccade latency at the SG location, the TL or a control location (with distractors appearing at all other locations). We calculated discrimination performance at these three different locations as a function of time. For pro-saccades, we replicated previous findings; discrimination performance at the TL/SG location improved the closer the DS onset was to saccade onset (55.6% at -300 ms (±40ms) to 71.1% at -100ms). For anti-saccades, we found that performance was best at the SG location (77.8%), however performance was also higher at the TL (54.7%) compared to control locations (38.7%). In terms of timing, similar to pro-saccades, performance improved the closer the DS onset was to the anti-saccade onset at the SG location (75% at -260 ms to 87.9% at -100ms). For the TL, performance was high when the DS onset was long before saccade onset and decreased close to saccade onset (63.3% at -260 ms to 47.8% at -100ms). These findings suggest that attention can be directed to multiple locations depending on the task but that close to saccade onset, attention is predominately directed to the saccade goal location.
Meeting abstract presented at VSS 2016