EXPERIMENTS USING TASKS other than reaction time (RT) have been reported which indicate that self-pacing facilitates performance. For example, self-paced letter sorting yields a higher output than paced (Conrad, 1960). Self-paced performance, as compared to paced performance, on the pursuit rotor yields higher scores, flatter performance curves, and a reduction in the between-subject differences in performance levels (Nance, 1960). The following two experiments were carried out to determine possible differences between paced and self-paced performance in a simple serial reaction task.
EXPERIMENT I
SubjectsSix paid female Ss, aged 25 to 35 years, were employed in the experiment. Five of the Ss had had at least ten sessions of paced RT in previous experiments and the other was given three daily sessions of RT practice.
ApparatusThe display consisted of two white stimulus lights, positioned side by side, beneath each of which were mounted a response key and a reset key. A red warning light (W) was mounted between the two stimulus lights. Although two stimulusresponse complexes were available, only one was used in any condition. One complex was used for the paced conditions and the other for the self-paced. The afterperiod, warning light duration, and foreperiod were controlled by Hunter timers. A Hunter Klockcounter measured RT in msec. During the self-paced conditions, reaction time and reset intervals were recorded on an Esterline-Angus oen recorder. To eliminate extraneous noise effects a background of continuous white noise of 67 db. (re 0.0002 dyne/cm 2 ) was presented through earphones during each condition.
ProcedureFigure 1 shows schematically the five conditions of Experiment I (1-5) and the six conditions of Experiment II (6-11).Following an earlier analysis of RT intervals (Foley & Dewis, 1960): T = foreperiod = interval from RT response to stimulus onset in no W conditions or W onset to stimulus onset interval in W conditions; t = sub-foreperiod = interval from reset response to stimulus onset; T 1 -afterperiod = interval from RT response to W onset; t 1 = sub-afterperiod = interval from reset response to W onset; reset interval = interval from RT response to reset response onset.