Address:
|
190 Dysart Road, Winnipeg, Manitoba, R3T 2N2, Canada 190 Dysart Road, Winnipeg, Manitoba, R3T 2N2, Canada
|
Abstract:
|
Researchers can adopt one of many different measures of central tendency to examine the effect of a treatment variable across groups. That is, methods for examining the effect of a treatment variable use least squares means, trimmed means, M-estimators, medians, etc., etc. As well, some methods begin with a preliminary test to determine the shapes of distributions prior to adopting a particular estimator of the "typical" score. In our paper we compared a number of recently developed methods with respect to their ability to control Type I errors and their power to detect effects when data were nonnormal, heterogeneous, and the design was unbalanced. We also examined type of trimming (symmetric or asymmetric), the benefit of bootstrapping, the use of two different transformations to eliminate the effects of skewness, and the use of a preliminary test for symmetry. Results indicate that one can indeed obtain robust tests for equality of the "typical" score across distributions with methods that were investigated. As well, power differences were present.
|