Human timing performance

If a subject is involved in a task requiring strict temporal control and the timing demands of the task are going to change, the allocation of attention is crucial. In each experiment, subjects were required to perform a series of taps so as to produce one goal time and then tap so as to produce the same or another goal time. Experiment 1 used a visual presentation of the stimuli. This presentation provided subjects with an implicit representation of the time intervals to be produced and an explicit representation of the serial position at which to switch from the first goal time to the second. The results indicated subjects had no problems switching from one goal time to the other at the correct serial position. However, tapping performance not only became more variable but performance virtually came to a halt when subjects changed from one goal time to another. One explanation is that subjects did not prepare for changes until the first interval following the required switch. A second is that the results were partly due to subjects trying to map the visual presentation of the stimuli to the times to be produced by tapping. Experiments 2 and 3 used an auditory presentation of the stimuli to address these alternatives. This presentation provided subjects with an explicit representation of the time intervals to be produced and an implicit representation of the serial position at which to switch from one goal time to the other. The results of Experiment 2 indicated subjects did not always switch at the correct serial position. In Experiment 2, the sequence could not be hierarchically organized. However, the tapping sequence used Experiment 3 was hierarchically organized and subjects were instructed to use a counting strategy to aid in correct parsing of the sequence. However, the effect of switching at the wrong serial position was still present. A model that relies on the intimate relationship of attention and timing control are presented. Further, the role of the representation of the task variables are addressed in relation to the parsing errors found in Experiments 2 and 3