Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Alfred B. Kristofferson


The purpose of the present investigation was to examine the nature of the timing mechanisms used by humans in two very distinct tasks and to determine whether there was any evidence to suggest that there was a common timekeeper. The duration discrimination experiment involved the presentation of a train of isochronously spaced auditory pulses with the interval between the last two pulses either slightly longer or shorter than the interval between any other two immediately successive pulses. The subject was instructed to make a judgment as to whether the final interval was long or short relative to the preceding intervals. This task was patterned after Kristofferson's (1980) many-to-few duration discrimination task which did not incorporate an explicit pulse train preceding the test stimulus. The second task, which was modelled after Wing (1973), involved two phases of repetitive finger tapping. In the first phase, a train of auditory pulses was presented to the subject and the subject was instructed to tap a Morse telegraph key in synchrony with the pulses. The second phase was a continuation of the first in the absence of any pulses. Subjects were instructed to continue tapping at the rate defined by the exogenous pulses in the synchronization phase. The main reason for using the pulse train duration discrimination method was to evaluate the role of using explicit standard in the context of duration discrimination. Using Kristofferson’s (1976) Real Time Criterion Model to obtain variance estimates of the internal timekeeper, it was possible to determine how those estimates changed with changes in the base duration. The function relating variance to base duration is flat over short base durations and rises in accord with Weber’s law over longer base duration. This is in contrast to earlier findings with the same subject and similar amounts of practice in which the implicit standard many-to-few method of duration discrimination was used. In that experiment a quantal step function emerged (Kristofferson, 1980). This difference implies that the nature of the stimulus plays an important role in the functioning of the internal timekeeper. More specifically, it appears that implicit standards elicit a quantal timekeeping mechanism whereas explicit standards elicit a quantal timekeeping mechanism. Explicit standards appear to elicit timekeeping mechanisms that obey Webster’s law. Webster’s law also characterizes the relationship between the variance and mean of the internal timekeeper in the continuation phase of the finger tapping experiment. After extensive and concentrated practice, it was shown that the above relationship is better described as being linear in timekeeper standard deviation than in timekeeper variance. Since no stepwise increments occurred in these functions, it seems unlikely that a quantal timekeeper is involved in the timing of interresponse intervals. Thus, the statistical principle of timing appears to apply to both explicit standard duration discrimination and the timing of repetitive finger taps in the form of a proportional standard deviation model. Although Weber’s law better characterizes the tapping functions general, one subject’s bioas corrected timekeeper variance versus mean function is very similar to his step function obtained in a many-to-few duration discrimination task. This similarity suggests that the motor timekeeper may have quantal characteristics under some circumstances and that the perceptual timekeeper may be related to the timekeeper involved in motor movement. A second reason for using the tapping paradigm was to evaluate Wing’s (1973) Two Process Model over a wider range of base temporal intervals. In this regard all of the predictions of the Two Process Model were upheld when it was applied to the short base temporal intervals (T less than or equal to 466 msec) but a major prediction of the model often was not borne out when the model was applied to longer base intervals (greater than or equal to 734 msec). It was concluded that it is appropriate to apply the model to evaluate tapping performance using base intervals in this shorter range, but that its application to base intervals in the upper range is probably inappropriate.

Although the synchronization phase of the tapping paradigm has been used in the past primarily to start and set the motor timekeeper for controling continuation phase tapping, it was here evaluated with the use of the Stimulus as Clock Model. This new model, which has its roots in the Two Process Model, evaluates synchronous interresponse interval tapping performance. The model accurately predicts one major characteristic of the timing of synchronization phase timing of interresponse intervals. However, its utility in terms of helping one understand the nature of the internal timekeeping process is limited.

The role of concentrated practice was examined in the context of both experimental paradigms. In both, practice steadily and significantly reduced variability in performance. Such a reduction played a major role in the shape and location of the variance functions and thus in estimates of various parameters.

Results of the two experiments are discussed with reference to previous investigations of response-stimulus synchronization with an attempt to integrate the vastly different results on a theoretical level. This discussion led to a new method by which to generate various parameters of the internal timekeeping mechanism.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Psychology Commons