A New Perspective On Sequential Testing Procedures In Canonical Analysis: A Monte Carlo Evaluation.

A Monte Carlo study was conducted in order to compare four testing procedures in establishing the number of nonzero population roots in a canonical analysis. Three of the methods investigated were sequential testing methods which have lately fallen under heavy criticism. The fourth method was a test advocated by Harris (1976) to be a more appropriate test. Results demonstrated the effectiveness of the sequential testing procedures and the relative ineffectiveness of the Harris method. The conclusion was that the criticisms aimed at the sequential testing for nonzero roots in canonical analysis are simply not well-founded.