Shared Control for Training in Virtual Environments: Learning Through Demonstration?

This paper presents an experiment to determine the underlying learning mechanism by which shared control with error reduction improves training effectiveness for dynamic manual control tasks in virtual environments. Specifically, the authors test the hypothesis that the learning mechanism of shared control with error reduction is through demonstration of the preferred strategy in the early phases of training. Three training protocols were utilized in this study and the experimental results indicate that all the benefits of shared control in training cannot be incorporated into strategy demonstration. The authors conclude that learning strategy alone is not the main mechanism of motor skill acquisition when shared control with error reduction is employed for training.