| Summary: | 碩士 === 國立中正大學 === 心理學研究所 === 92 === The aim of this study is to discuss that the causation of switch cost is due to the additional processes or the proactive interference coming from the previous task. We measured the ERP components, P300、S-LRP and R-LRP, to examine the following three models of task switching. The fist model proposed by Merian (2000) suggest that switch cost happens in the stimulus identification stage which is related on P300 component. The second model proposed by Rubinstein et al. (2001) assume that there are two additional stages, goal-shifting and rule-activation, besides the task process, influence the stage of response selection. The third model proposed by Sohn & Anderson (2001) support the theory of proactive interference and assert that switch cost happens in the stage of response selection. A critical issue to disentangle the three models is to examine whether the effects of task switching or task cueing will happen in the stage before response selection, and whether the two effects are independent or interactive.
In this experiment, we increased the number of S-R mappings and manipulated whether there is a cue provided before the task stimulus onset or not, and whether the task of the present trial is the same with the previous one or not. The results of this experiment show that there is a significant main effect of task-cueing reflected on the peak latency and the peak amplitude of the P300 component. There are significant main effects of task switching and task cueing but no significant interaction on stimulus-locked LRP interval. The results of this research reconcile best with the carryover model proposed by Sohn & Anderson (2001).
|
|---|
