Task switching and subjective duration.

Three experiments investigated the effects of switching from a nontiming task (addition of rapidly presented digits) to estimation of the duration of short tones or visual stimuli (Experiments 1 and 2), or the production of time intervals (Experiment 3). In general, compared with trials without a task-switch, trials involving a switch resulted in shorter duration estimates, but longer productions. The difficulty of the nontiming task, and the gap between the nontiming task and the timing task, also played a role, at least in some cases. Verbal estimates were shortened more at longer stimulus durations than at shorter ones in both experiments using this method. In general, the task-switch procedure produced opposite effects to "speeding up the clock" manipulations. The data were discussed in terms of models involving internal clock switch processes, including the notions of an "attentional gate" or a "flickering switch".