Contextual interference effect in motor skill learning : examination of attention demands

Abstract

The contextual interference (Cl) effect refers to the important differences observed between high Cl (e.g., random schedule) and low Cl (e.g., blocked schedule) practice in motor skill learning, in which acquisition performance suffers but retention performance is facilitated by high Cl practice. Explanations based on forgetting and elaborative processing both intimate that the additional cognitive activity required during high Cl practice contributes to the emergence of this effect. The purpose of the present study was to investigate the attention demands associated with the two practice conditions by using a dual-task method. The primary task was to learn three key-pressing patterns with three segmental timing requirements for each pattern in either random or blocked practice schedule. In Experiment 1, an auditory stimulus (i.e., high or low tone) was presented unpredictably during the practice trials. Subjects were required to concentrate on the primary task during the acquisition while also monitoring the number of the high tones. The acquisition, which consisted of three blocks of 18 trials on the timing task, was followed by 10-min and 24-hour delayed retention tests. Results showed the typical Cl effect with retention performance being facilitated by the random practice, even though acquisition performance was superior for the blocked practice. Random practice subjects produced more errors on the secondary task, indicating that greater attentional demands were required during random practice. However, a cautious interpretation of these data was necessary because degraded primary task performance was observed for the dual-task condition, compared with the control condition. Experiment 2 attempted to eliminate this problem, but failed to replicate the findings on the secondary task performance. Using a discrete reaction time (RT) probe task in Experiment 3, the findings from Experiment 1 were replicated, as well as eliminating the primary task problem observed in Experiment 1. The data from Experiment 3 are consistent with the predictions from the elaboration view. In Experiment 4, a modified practice condition was employed to directly test the forgetting hypothesis. The results provide support for the contribution of a reconstructive process to random practice. These data are discussed with respect to the contribution of forgetting and elaboration explanations for the Cl effect.