Paced and self-paced continuous reaction time

Studies of the effect of pacing in continuous RT tasks are inconclusive. They have failed to (1) compare paced and self-paced RT at matched error rates and (2) examine RT distributions as well as mean or median RT. In the present experiment RTs were corrected for error rate by relating them to the speed-accuracy tradeoff function, which was measured in each experimental session. In this way deciles of the RT distributions of paced and self-paced conditions were compared, for eight subjects and four successive sessions. The task was a four-choice continuous RT task, employing visual stimuli. The results show that RT distributions are wider in the paced conditions. RTs of the first deciles are faster for pacing whereas medians in paced and self-paced conditions are about equal. The results can be explained when it is assumed that the variable R-S intervals in paced conditions lead to a larger variability of RT.