ABSTRACT
This chapter discusses why heterogeneity of variance is a reasonable expectation in many training evaluation situations, and how that heterogeneity can affect testing for mean differences. Although the literature dealing with training effects on variability is somewhat limited, several individuals have made important contributions to the beginnings of a theory in research area. The chapter examines that the standard deviation or its squarevarianceis the most common metric for variability, and thus has focused on measure. The chapter focuses on the potential importance of variability effects for organizational training, both through the possible impact of such effects on traditional approaches to training evaluation and through the development and testing of training variance effects in their own right. An organization wants to test the effectiveness of a stress-reduction program. A five-factor model is proposed for the prediction of variability effects due to training, and research propositions based on the model are advanced.
