Summary and frequency of KR presentation effects on retention of a motor skill

This experiment investigates the recent proposal of Schmidt, Young, Swinnen, and Shapiro (1989) that summary knowledge of results (KR) improves skill retention. In Schmidt et al.'s experiment, however, summary length varied concomitantly with the frequency of KR presentation. The current investigation held KR presentation frequency constant while manipulating the number of trials seen in the summary KR display. Subjects were required to perform a timing task on a linear slide. Five groups (n = 12) of subjects participated in acquisition trials then in 10-min and 2-day delayed no-KR retention tests. In 4 conditions, subjects completed each acquisition block without any KR, but following each block they received KR on either 15, 7, 3, or 1 of the 15 trials performed in that block. In the final condition subjects received immediate KR. Analysis of the absolute constant error (magnitude of CE) data for acquisition revealed all groups improved with practice and the immediate KR group performed better than all the summary groups which in turn did not differ significantly. Analysis of the magnitude of CE retention data found performance to be worse on the 2-day retention test for all groups. The effect of condition was significant. The 1/1 group had lower error scores than all other groups, which in turn were not significantly different. Analyses of variable error (VE) revealed only that VE decreased with practice. These findings suggest frequency of KR presentation may be the basis for the summary KR effect found by Schmidt et al. (1989).