Using Van Valens Procedure In Business Research To Assess Consistent Differences In Multidimensional Variability In Two Or More Groups

Abstract

Much business research involves comparisons in two or more groups on many dimensions.  This paper primarily focuses on demonstrating and providing guidance as to how researchers should approach a multivariate analysis in the comparison of sets of corresponding characteristics in two or more independent groups.  In particular, this paper demonstrates the utility of a simple but not widely known procedure developed by Van Valen (1978) that should be employed to test for the significance of differences in overall variability in the sets of corresponding characteristics in two or more groups, a test that enjoys much statistical power in detecting significant subtle group differences when the set of characteristics in one group consistently demonstrates greater variability than the corresponding set of characteristics in the other group(s).