Diminishing the effects of reduced frequency of knowledge of results on generalized motor program learning

Several recent studies have shown that, compared with presenting knowledge of results (KR) on every practice trial, withholding KR on some practice trials enhances the learning of generalized motor programs (GMPs; Wulf, Lee, & Schmidt, 1994; Wulf & Schmidt, 1989; Wulf, Schmidt, & Deubel, 1993). In this study, whether this effect may result from an uncertainty on the part of participants about when KR is to be presented was determined by examining the GMP learning of two 67% relative-frequency KR groups-one that received advance information regarding the delivery of each KR (67% AKR) during practice trials and another that received no advance information (67% KR)-as well as that of a 100% KR group. The task required participants to produce 3 movement patterns that shared the same relative and absolute timing and relative amplitude but differed in terms of absolute amplitude. KR was provided by displaying the root-mean-square error (RMSE) score and by graphically superimposing the participant-produced pattern on that of the goal movement. The results revealed no group differences in measures of GMP development or parameterization effectiveness during practice and no-KR retention. However, during no-KR transfer with a novel absolute amplitude, the 67% KR group demonstrated a more accurate and stable GMP than the 67% AKR and 100% KR groups. Possible explanations for why advance knowledge about KR delivery diminishes GMP development are discussed.