Temporal control of rhythmic performance: a comparison between young and old adults.

Young (20-30-year-old) and older (60-76-year-old) adults were tested on two measures of rhythmic performance. The first involved tapping at the subject's own preferred rate, a measure of so-called internal tempo. Over five sessions of testing, tapping rates were consistently and significantly slower on average in the older subjects than the younger ones, but rates were not relatively more variable in older subjects (i.e., coefficients of variation, standard deviation/mean, did not differ between the older and young people). In addition, both old and younger subjects performed on a synchronized-tapping and continuation task of the type used by Wing and Kristofferson (1973, Perception and Psychophysics, 14, 3-12). Target interresponse times were 300, 400, 500, 600, and 700 ms, and in all cases interresponse intervals produced by both the old and young adults matched the target times very closely. Wing and Kristofferson's analytical procedure was used to decompose tapping variance into that attributable to timing processes and that resulting from motor implementation of the timing signal. Both sorts of variance increased with increasing target interresponse time (with timer variance increasing most markedly), but no difference was found in either type of variance in comparisons between the old and younger subjects. If the internal tempo measure directly reflects the speed of internal timing processes, the data suggest that such processes are slower, but not relatively more variable, in older than younger subjects (consistent with some previous evidence and speculation), but that the calibration of performance forced by the synchronization task will make such an age-related difference in "internal clock speed" unobservable on synchronized-tapping tasks.