Effects Of Pre - Stimulus Consistency And Task Preparation On Response Repetition

In task switching paradigm, response repetition produce benefits (RR benefit) if the task repeats, but cost (RR cost) if the task switches. The present study made use of Event-related potentials (ERPs), employing a sequential two-task procedure, to investigate how previous-trial congruency and task preparation affect the response repetition effects (RR effects).In task1, we manipulated the previous-trial congruency with univalent-congruent S1and univalent-incongruent S1stimuli that consisted of three items of the same item category (e.g. congruent:AbA/incongruent:AEA) in experiment1, but with bivalent-congruent and bivalent-incongruent stimuli (e.g. congruent:A5A/incongruent:b4b) in experiment2. Task2in both experiments used neutral S2stimuli (e.g.*6*/*E*). The central item was always the target, and participants were asked to make a judgment whether the letter is upper or lower case, or whether the number is odd or even. The behavioral data indicated that, in task1, the response time was longer to incongruent stimuli than to congruent stimuli, and the accurency for congruent trials are higer than for incogrent trials. The reaction time and congruency in two experiments were not mediated by the valence varible. RR cost was larger in task shift sequence than task repeat sequence. RR cost did not show any difference between with or without task preparation when task shifts, and when task repeat, RR benefit was larger in trials without a task cue the onset of S2than trials with a task cue before S2. Previous-trial congruency has no effect on RR cost in task shift sequence. We found salient category priming effect in electrode FZ, the ERP elicited by RR sequence was more positive than RS sequence when task repeated; and also found response inhibition in both task sequences from stimulus-locked lateralized readiness potentials (LRP) after S2; but stimulus-locked LRP before the S2revealed no response inhibition happened when the stimuli in task1were univalent. In experiment2, we observed response inhibition and the effect of previous-trial congruency on the response inhibition before S2, with larger inhibition after responding to congruent trials than incongruent trials. As to the task preparation, it has no salient effect on the response inhibition before the onset of stimulus2, but the increasement of response inhibition after the onset of S2was sped up with task preparation. These findings indicate that, there are two machinism contribute to the RR effects, including response inhibition and category priming. But the effect of previous-trial congruency on strength of response inhibition probably does not depend on the strength responding to S1; the task preparation may be not necessary to inhibition, but affect the speed of inhibition increasement after the onset of S2.