Substituting some unassisted practice with robotic guidance: Extending support for mixed practice within a rhythmic sequential task


There is equivocal evidence regarding the effectiveness of robotic guidance on (re)learning of voluntary motor skills. Robotic guidance can improve the performance of continuous/ tracking skills (steering: Marchal-Crespo et al., 2010; tracing: Williams et al., 2016), although seldom more effective than unassisted practice alone (e.g., pointing: Manson et al., 2014; trajectory tracing: Liu et al., 2006). However, previous studies employed robotic guidance on all trials. Recently, we showed that mixing robotic guidance with unassisted practice (i.e., mixed practice) can significantly improve the learning of a golf putting task (Bested et al., 2019a & b). Further, studies involved self-paced movements, thus less applicable in some daily activities. As such, the current study aimed to investigate the influence of mixed practice on the timing accuracy of a rhythmic, sequential task. Two groups of participants performed circle-drawing sequences in synchrony to rhythmic auditory signals. They completed a pre-test and an acquisition phase, followed by immediate retention and transfer post-tests. One group received robotic guidance on 50% of the acquisition trials (i.e., mixed practice) whereas another group always practiced unassisted. The pre-test, post-test, and transfer test were performed unassisted. Both groups exhibited significant improvements in their timing accuracy between the pre-test and the immediate retention and transfer post-tests. This study provides further evidence that mixed practice (i.e., substituting some unassisted practice with robotic guidance) can reduce the physical demands and facilitate the (re)learning of voluntary actions (e.g., rehabilitation, music, sports).

Acknowledgments: Natural Sciences and Engineering Research Council of Canada (NSERC), Ontario Research Fund (ORF), Canada Foundation for Innovation (CFI)