×

Type I error control of two-group multivariate tests on means under conditions of heterogeneous correlation structure and varied multivariate distributions. (English) Zbl 1079.62517

Summary: A Monte Carlo study was conducted to evaluate the performance of two-group multivariate tests on means under conditions of homogeneous and heterogeneous correlation structure and multivariate normality and nonnormality. The test statistics under study included those based on the work of Hotelling and alternative procedures proposed by James, Yao, Johansen, Nel and van der Merwe, and Kim. Results indicated that the F statistic based on Hotelling’s \(T^2\) was outperformed by the alternative procedures which showed good Type I error control under moderate to large sample sizes. Under small sample sizes, the alternative procedures tended to have unacceptable Type I error control. The F statistic based on Hotelling’s \(T^2\) is recommended for use when the sample size in each group is small (e.g., \(n = p + 1\)) and correlation heterogeneity is mild \((<.3)\), or when groups are approximately equal in size. Though simulated conditions of multivariate nonnormality had some impact on the Type I error rates on some of the procedures, the impact was in general small and apparent only under the smaller sample size conditions.

MSC:

62H15 Hypothesis testing in multivariate analysis
65C05 Monte Carlo methods
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] James G.S., Biometrika 41 pp 19– (1954)
[2] Yao Y., Biometrika 52 pp 139– (1965)
[3] DOI: 10.1093/biomet/67.1.85 · Zbl 0422.62066 · doi:10.1093/biomet/67.1.85
[4] Nel D.G., Communications in Statistics–Simulation and Computation 15 pp 3719– (1986)
[5] DOI: 10.1093/biomet/79.1.171 · Zbl 0850.62427 · doi:10.1093/biomet/79.1.171
[6] DOI: 10.1037/0033-2909.108.2.308 · doi:10.1037/0033-2909.108.2.308
[7] DOI: 10.2307/1165116 · doi:10.2307/1165116
[8] DOI: 10.2307/1164656 · doi:10.2307/1164656
[9] DOI: 10.1080/03610919708813439 · Zbl 1100.62575 · doi:10.1080/03610919708813439
[10] de la Rey N., South African Statistical Journal 27 pp 129– (1993)
[11] DOI: 10.2307/2286719 · Zbl 0398.62029 · doi:10.2307/2286719
[12] DOI: 10.1037/0033-2909.86.6.1255 · doi:10.1037/0033-2909.86.6.1255
[13] DOI: 10.2307/2282915 · doi:10.2307/2282915
[14] DOI: 10.2307/2283331 · doi:10.2307/2283331
[15] DOI: 10.1093/biomet/58.1.105 · Zbl 0218.62081 · doi:10.1093/biomet/58.1.105
[16] DOI: 10.1093/biomet/60.1.107 · Zbl 0254.62041 · doi:10.1093/biomet/60.1.107
[17] Coombs W.T., Review of Educational Research 66 pp 137– (1996) · doi:10.3102/00346543066002137
[18] Hotelling H., Proceedings of the Second Berkeley Symposium on Mathematical Statistics and Probability (1951)
[19] Beasley, T.M. and Sheehan, J.K. 1994. Choosing a MANOVA Test Statistic When Covariances are Unequal. Paper Presented at the Annual Meeting of the Midwestern Educational Research Association. October1994, Chicago, IL. ERIC 379296
[20] DOI: 10.1007/BF02919505 · Zbl 0042.38103 · doi:10.1007/BF02919505
[21] DOI: 10.1007/BF02293811 · Zbl 0388.62023 · doi:10.1007/BF02293811
[22] DOI: 10.1007/BF02293687 · Zbl 0521.65003 · doi:10.1007/BF02293687
[23] Bradley J.V., British Journal of Mathematical and Statistical Psychology 34 pp 144– (1978) · doi:10.1111/j.2044-8317.1978.tb00581.x
[24] DOI: 10.1111/j.2044-8317.1992.tb00993.x · doi:10.1111/j.2044-8317.1992.tb00993.x
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.