The effect of withdrawal of knowledge of results on performance of movement at different speeds

The present study addressed the extent to which KR, its initial presence and later withdrawal, influenced the learning of a motor skill performed at different speeds.

Forty eight right handed female students served as subjects for this study. Prior to the first day of testing, each subject was randomly assigned to one of four movement speed conditions. All subjects attempted to produce a 70 centimeter movement from a start button to a padded target. The four criterion times were established at 250, 500, l000, and 2o0o milliseconds, thus producing respective speeds of 280, 140, 70 and 35 cm/second. Al1 subjects received 5 practice trials (Trials 1-5) with KR (error and direction in milliseconds), 40 trials (Trials 6-45) with KR, and then 40 trials (Trials 45-85) without KR.

Deviations from the criterion target were recorded (with respect to sign) to the nearest millisecond. The raw movement time data were converted into three dependent variables: (l) absolute error (AE), (2) constant error (CE), and (3) variable error (VE) scores. Analysis of variance was used to analyze the data. A Newman-Keuls post hoc test was administered where appropriate to examine all significant differences.

iii

iv The results indicated that during practice with KR, performance accuracy (AE) and consistency (VE) significantly improved for all speed groups (p < .0l). When KR was withdrawn consistency remained fairly constant however accuracy diminished significantly for all speed groups (p < .0l) which indicated that subjects were unable to improve their performance without KR.. The results also revealed that the faster movements, those at movement times of 250 and 500 milliseconds with respective speeds of 280 cm/second and l40 cm/second, were more accurate and more consistent than were the groups with slower criterion movement times of l000 and 2000 milliseconds with speeds of 70 cm/second and 35 cm/second, respectively. From this analysis of CE, the results indicated that subjects with the slowest movement time criterion, 20oo milliseconds, with a speed of 35 cm/second, had the most erratic periods of bias, moving from the greatest extremes early to the greatest extremes late when KR was given and performance remained almost as erratic when KR was withdrawn.