Acquisition and Extinction of a Simple Motor Skill as a Function of Delay of Knowledge of Results

Abstract

The present experiment demonstrates that the effects of delay of knowledge of results (KR) in a line drawing task depend upon the dependent variable which is considered and the original response tendency of the subjects. Delay of KR interfered with the acquisition of the correct response when number of correct responses is the dependent variable. When KR was omitted the immediate-KR group continued to make more correct responses than the delayed-KR group. However, there was a significant reduction in correct responses for both groups. When absolute error was the response measure there were no significant differences between immediate-and delayed-KR groups either during acquisition or extinction. Analysis of the type of response made during extinction suggested that the overshooting effect obtained by previous investigators may be typical of short responders trained under conditions of immediate reinforcement but not of those trained under delayed-KR. Greenspoon and Foreman (1956) obtained clear cut evidence that the effectiveness of knowledge of results (KR) in a line drawing task varies inversely with the time delay between the response and KR. On the other hand, Bilodeau and Bilodeau (1958) were unable to demonstrate a similar effect in five different experiments. Bilodeau and Ryan (1960) hypothesized that the Greenspoon and Foreman findings could not be replicated and they did in fact obtain null results. However, Dyal (1965) has recently replicated the Greenspoon and Foreman results for the case of a 30 sec. delay interval. The purpose of the present experiment is twofold: (a) to provide replication of Green-spoon and Foreman's study at the 20 sec. delay interval used by Bilodeau and Ryan, and (b) to determine the effect of elimination of KR on a simple motor response which has been formed on the basis of KR.

Effects of delay of knowledge of results and subject response bias on extinction of a simple motor skill.

“…Two previous experiments in this series (Dyal, 1964;Dyal, Wilson, & Berry, 1965) have supported the findings of Greenspoon and Foreman (1956) that the operation of delaying knowledge of results (KR) tends to interfere with the acquisition of a linedrawing skill. 2 Furthermore, in the Dyal, Wilson, and Berry experiment it was observed that within that group of .S' s whose original response bias was to undershoot the target (short respond-1 The data for this experiment were collected by Gene Arrington.…”

Self-controlled feedback is effective if it is based on the learner’s performance: a replication and extension of Chiviacowsky and Wulf (2005)

“…This marginally greater delay between movement completion and KR delivery could be argued to have allowed the engagement of additional error estimation processes to benefit learning that were not available in the fixed 2000 ms interval. However, we specifically adopted a 2000 ms KR delay interval based on past research revealing that error estimation processes are engaged immediately following a movement ( McGuigan, 1959 ; McGuigan et al, 1960 ; Dyal et al, 1965 ; Dyal, 1966 ; Newell, 1976 ; see Salmoni et al, 1984 ; Swinnen, 1988 ; Swinnen et al, 1990 for in-depth discussions). Therefore, any error estimation processes would be expected to have occurred very quickly following movement completion and well within the fixed 2000 ms KR delay interval used for all groups.…”

Knowledge of results and motor learning: A review and critical reappraisal.

“…But Swinnen et al (1984) found that immediate KR degraded learning relative to a condition with an 8-s delay (confounded with post-KR delay). Dyal et al (1965), Dyal (1966), andMcGuigan et al (1960) with line drawing, Schmidt and Shea (1976) with a rapid positioning task, Schmidt et al (1975) with rapid timing, and Boulter (1964) with slow positioning failed to find any effect of KR delay (confounded with post-KR delay). However, Dyal (1966) showed that subjects with different response biases in acquisition performed differently on a no-KR transfer test as a function of KR delay in acquisition, suggesting that KR delay affected learning in some way.…”

“…When KR delay is varied and the post-KR interval is held constant (KR delay confounded with intertrial interval), Bourne and Bunderson (1963), with a concept-formation task, Boucher (1974), Koch and Dorfman (1979; rapid movement), Marteniuk (1981), Mc-Guigan (1959), and Salt/man, Kanfer, and Greenspoon (1955) failed to find reliable effects on performance. When KR delay is varied and the intertrial interval is held constant (KR delay confounded with post-KR delay), Becker, Mussina, and Persons (1963), Bilodeau and Ryan (1960), Boulter (1964), Dyal (1966), Dyal, Wilson, & Berry (1965), Larre (1961), Noble and Alcock (1958), Ryan and Bilodeau (1962), Schmidt and Shea (1976), Schmidt, Christenson, and Rogers (1975;rapid task), Swinnen, Schmidt, and Shapiro (1984), and Timmons and Wiegand (1982;simple task) found no effects of KR delay on performance. Archer and Namikas (1958), using the delay of a tone indicating that the subject had contacted the target during a tracking-task trial, also found no effect.…”