Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Richard A. Magill


Four experiments examined the generalizability of the contextual interference (CI) effect. In all experiments, to examine the locus of the CI effect measures assessing generalized motor program (GMP) learning and parameter learning were used in addition to a measure of general performance. In the first two experiments, variations of a timing tapping task, which were controlled by either different GMPs (Experiment 1) or the same GMP (Experiment 2), were learned under either blocked (low CI) or serial (high CI) practice. The CI effect was found for the general performance measure regardless of the task characteristics, contrary to Magill and Hall's (1990) hypothesis that tasks controlled by the same GMP do not create the CI effect. The dissociated measures of GMP learning and parameter learning showed that parameter learning was enhanced by high CI practice in both experiments and there was a tendency that GMP learning was also enhanced by high CI practice in Experiment 1. To extend the findings with timing characteristics, Experiment 3 involved task variations requiring modifications of the overall force parameter of the same GMP. Consistent with the results of Experiment 2, the results showed the CI effect for overall force parameter learning that was also reflected in the general performance. In Experiment 4, task variations requiring simultaneous modifications of both overall duration and overall force parameters of the same GMP were learned. The results showed the CI effects for both types of parameter learning, which were reflected in the general performance. Different amounts of practice used in Experiments 3 and 4 did not influence the efficacy of the CI effect. Findings based on retention and transfer tests in Experiment 4 were compatible with each other. Thus, the CI effect was found regardless of the task characteristics, the number and types of parameters modified, the amount of practice, and types of learning tests. These results indicate that the CI effect is generalizable to more various motor learning situations than previously believed because any aspect of performance that changes from trial to trial during practice leads to better retention and transfer regardless of the amount of practice.