Switching between simple response-sets: inferences from the lateralized readiness potential.

Behavioral studies have documented that task switching incurs a longer reaction time than task repetition, and that advance cueing information about the forthcoming task reduces mean reaction time. The present study used P300 peak latency and two lateralized readiness potential (LRP) intervals--stimulus-locked and response-locked--to infer the loci of task switch and task-cueing effects and how they may interact in the basic task processing chain. Participants performed two tasks in a random order, so that on each trial they either repeated the task from the previous trial or switched to another task. In one condition, each stimulus was preceded by a cue informing participants which of the two tasks to perform; and in the other condition, each stimulus was preceded by a non-informative cue. Results indicated that both mean reaction times and the stimulus-locked LRP intervals were longer for switch than repeated trials, whereas P300 peak latencies and response-locked LRP intervals were identical for both trials. Similarly, both reaction times and the stimulus-locked LRP intervals were longer for no task-cueing than for task-cueing conditions, and P300 peak latencies and the response-locked LRP intervals were identical for both conditions. Finally, task switch and task-cueing effects appeared to be approximately additive, indicating the two factors influence distinct stage processes. We suggest that task switching resulted in prolongation of the response selection process by carry-over priming effects from the previous task, whereas task-cueing shortened the duration of the earlier process before response selection on both switch and repeated trials.