Canonical Variate Analysis (CVA) has two major problems: sampling error and accumulated rounding error. This paper describes a method designed to minimize these problems. The combined effect of the two problems was assessed for CVA and the new technique in an empirical study. The new technique appears to be superior to CVA when the ratio of variables to sampling units is greater than 1/10.
No takes yet. Share an insight, caveat, or question.
Weinberg et al. (1976) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: