Evaluation of Motor Learning When Performance Levels are Heterogeneous

Abstract

In three motor learning experiments (jumping, speed of arm movement and balancing), raw learning scores are unrelated to final degree of skill but are negatively correlated with initial performance. Using the method of residuals to hold statistically constant the influence of initial skill, it is possible to compute derived learning scores which are unrelated to initial skill and at the same time substantially correlated with learning and also with post-learning performance. Since these scores estimate the individual learning that would have occurred if each person had started at the same level of initial skill, they should prove practically useful as true measures of motor educability. The method can readily be applied to the construction of tabular norms in which several variables must be taken into account. It has also been used to define an endurance index.