Summary knowledge of results (KR) involves delaying the presentation of KR until a predetermined number of trials has been completed. Schmidt, Young, Swinnen, and Shapiro (1989) found an optimal summary length of 15 trials (compared to 1, 5, and 10) and Schmidt, Lange, and Young (1990) found an optimum length of 5 trials (compared to 1, 10, and 15) for 1 and 2 kinematic degree of freedom timing tasks, respectively. Experiment 1 was designed to determine the optimal summary length for learning a simple isometric force production task. The results indicated better retention for the 16-trial summary than for 8- or 1-trial summaries. Experiment 2 was an initial attempt to determine the locus of the 16-trial summary effect. Specifically, Experiment 2 focused on the role played by both the information provided in the summary presentations and the practice per se of the first 14 trials in the 16-trial block. The results suggest KR is important to ensure response stability. However, it also appears important to restrict the subject's immediate utilization of KR. This latter requirement may potentially aid the parameterization process, resulting in a reduction in response bias.
