Convergent and discriminant relationships between factor matrices of motor performance and arousal data.

An experimental study was conducted to demonstrate the value of a procedure for analyzing motor performance data. A sample of 156 subjects practiced on two instruments for motor learning while arousal data were recorded. Each set of data was submitted to an analysis of principal components and four components resulted for each set of data. Similarity coefficients were calculated for pairs of component matrices after rotation to maximum similarity. The similarity coefficients exhibit a consistent pattern which provides meaningful information concerning influence of experimental conditions on performance data.