Enhancing Motor Learning through Theoretical and Experimental Methods Acquisition, Retention, and Learning Strategies

The first goal of this thesis is to apply theoretical and experimental methods to investigate the mechanisms of human motor learning. By comparing behavioral results with theoretical predictions, the second goal is to identify components of learning that can be further improved. Specifically, the acquisition, retention, and self-directed strategies of motor learning were studied. Previous learning studies have shown that spacing learning sessions in time increases the efficacy of verbal skill acquisition. I showed that an increased rate of acquisition in motor adaptation was also observed when the temporal separation of training trials was increased by a few seconds. I hypothesized that each instance of training movement spawns a memory trace of the experience from which the learner continues to sample until another movement was made. I showed that indeed the learner's sensitivity to error (trial-to-trial learning rate) increases with the rest period after a movement with a time constant of about 4 seconds. An alternative theory hypothesized that time passage increases the learner's uncertainty about the task, and the trial-to-trial learning rate should increase with the uncertainty. The fact that the trial-to-trial learning rate was a monotonically increasing function of post-movement time contradicted the pre-movement time dependency as predicted the uncertainty-based model. In learning a motor task, a learner does not only need to estimate the optimal parameters for that task, but also learn the structure of the task itself. However, most motor learning literatures do not vary the task structure. Secondly, learning task structure often takes place at the same time as task parameter estimation. I introduced a novel iii paradigm that varied the task structure: adaptation to a gradually or a suddenly introduced dynamical perturbation of the same magnitude. By measuring the passive decay of adaptation in naïve and experienced learners, the manifestation of structural learning independent of parameter estimation could be observed. I showed that gradual training promoted retention of adaptation in naïve but not necessarily experienced learners. However, the gradualness alone was sufficient to promote retention in retraining sessions suggesting that a prior of the task structure was acquired at the first exposure and continued to influence people's motor adaptation and retention properties in subsequent learning sessions. In real life, we are often faced with the problem of having to learn by ourselves. I explored the interactive component of motor learning and the optimality of the strategy people use in selecting the sequence of movements …