NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED360348
Record Type: Non-Journal
Publication Date: 1993-Apr
Pages: 19
Abstractor: N/A
Reference Count: N/A
ISBN: N/A
ISSN: N/A
The Effect of the Multivariate Box-Cox Transformation on the Power of MANOVA.
Kirisci, Levent; Hsu, Tse-Chi
Most of the multivariate statistical techniques rely on the assumption of multivariate normality. The effects of non-normality on multivariate tests are assumed to be negligible when variance-covariance matrices and sample sizes are equal. Therefore, in practice, investigators do not usually attempt to remove non-normality. In this simulation study, the effects of non-normality on skewed multivariate data in terms of power were examined by manipulating the factors such as distribution, sample size, number of variables, and variance-covariance matrix. The number of replications was set to 500, and sample sizes of 10, 15, and 20 were used, with 2 sets of variables, and 2 variance-covariance matrices. The multivariate Box-Cox transformation was applied to remove non-normality. The power of multivariate analysis of variance (MANOVA) was then calculated after the transformation. The results were compared with the power calculated before the multivariate Box-Cox transformation was applied. In conclusion, even when variance-covariance matrices and sample sizes were equal, small to moderate increases in power were observed. (Author/SLD)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A