Too Much of a Good Thing: Random Practice Scheduling and Self-Control of Feedback Lead to Unique but Not Additive Learning Benefits

We examined the impact of self-controlled knowledge of results on the acquisition, retention, and transfer of anticipation timing skill as a function of random and blocked practice schedules. Forty-eight undergraduate students were divided into experimental groups that practiced under varying combinations of random or blocked as well as self-controlled or yoked practice conditions. Anticipation timing performance (5, 13, and 21 mph) was recorded during acquisition and during a short term no-feedback retention test. A transfer test, administered 24 h after the retention test, consisted of two novel anticipation timing speeds (9, 17 mph). Absolute error (AE) and variable error (VE) of timing served as the dependent measures. All participants improved their accuracy and consistency across acquisition blocks; however, those who practiced under blocked rather than random conditions had greater accuracy (lower AE) regardless of feedback delivery. During retention and transfer, those who practiced under random conditions showed greater consistency (lower VE) compared to their blocked counterparts. Finally, participants who controlled their feedback schedule were more accurate (lower AE) and less variable (lower VE) during transfer compared to yoked participants, regardless of practice scheduling. Our findings indicate that practicing under a random schedule improves retention and transfer consistency, while self-control of feedback is advantageous to both the accuracy and consistency with which anticipation timing skill transfers to novel task demands. The combination of these learning manipulations, however, does not improve skill retention or transfer above and beyond their orthogonal effects.