Abstract
Event-related potentials (ERPs) are usually used to study visual attention in the absence of eye-movements. For example, in visual search, about 200 ms after stimulus onset, more negative potentials are observed at posterior electrodes contra- than ipsilateral to a searched-for stimulus (N2pc or PCN). It is usually assumed that such laterality effects, as the N2pc, indicate attentional selection of the target stimulus before saccade execution, although this has not been investigated without constraining saccade parameters in some way either during the task, or later in the analysis. Here, we present an approach to investigate pre-saccadic attention shifts in an unconstrained way. Participants searched for a color-defined target in single-target and two-targets blocks while ERPs and eye-movements were recorded. Single-target blocks required a single saccade; two-targets blocks required either a single saccade to one of the targets or two successive saccades to both targets. In two-targets blocks, targets could appear on the same or on opposite sides of the vertical midline. If targets captured attention and pre-saccadic attention shifts to saccade target locations were necessary for saccade execution, we would expect enhanced attentional competition between (1) two targets compared to single targets; (2) two opposite-sides targets compared to two same-side targets; and (3) two saccades rather than one saccade conditions. More attentional competition was expected to delay saccade latency and to weaken and delay pre-saccadic laterality effects in ERPs. By means of temporal alignment we time-locked ERPs simultaneously to stimulus onsets, saccade onsets, and saccade offsets. Predictions (1) and (2) were partly and fully confirmed, respectively, but no evidence was found for (3). We explain the implications of our results for the role of attention during saccade preparation, and we point out how our method of temporally aligning ERPs compares to ICA-based EEG artifact correction procedures and to psychophysical dual-task approaches.
Meeting abstract presented at VSS 2016