Main Article Content
The objective of this research is to compare the efficiency between two test statistics: Li and Chen’s test statistic and Srivastava et al.’s test statistic, for testing the equality of two covariance matrices for high-dimensional data distributed as multivariate normal (p variables). There are two criteria for comparing the tests consisting of the probability of type I error and power of the test which were measured through data simulation using Monte Carlo technique 1000 iterations. The two high-dimensional data distributed as multivariate normal data (p variables) under five covariance matrix structures: compound symmetry, simple, Toeplitz, unstructured, and variance components, were simulated. The number of variables (p) was assigned to be greater than or equal to its sample sizes and varied in . It was shown that Li and Chen’s test statistic can control the probability of type I error for all covariance matrix structures considered whereas Srivastava et al. is unable. In addition, the Li and Chen’s test statistic was strongly higher power than the other one and its power converged to one when p and sample sizes were increased for all situations.
Keywords: multivariate normal distribution; probability of type I error; power of the test; covariance matrix
 กัลยา วานิชย์บัญชา, 2554, การวิเคราะห์ข้อมูลหลายตัวแปร, ภาควิชาสถิติ คณะพานิชยศาสตร์และการบัญชี จุฬาลงกรณ์มหาวิทยาลัย, กรุงเทพ, 589 น.
 Anderson, T.W., 2003, An Introduction to Mutivariate Statistical Analysis, 3rd Ed., Wiley-Interscience, New york, pp. 265-267.
 Schott, J. R., 2007, A high-dimensional test for the equality of the smallest eigenvalues of a covariance matrix, J. Multivariate Anal. 97: 827-843.
 Srivastava, M.S., 2007, Testing the equality of two covariance matrix and testing the independence of two subvectors with fewer observations than the dimension, Proceedings of the International Conference on Advanced in Inter-disciplinary Statistics and Combinatorics.
 Srivastava, M.S. and Yanagihara, H., 2010, Testing the equality of several covariance matrices with fewer observations than the dimension, J. Multivariate Anal. 101: 1319-1329.
 Li, J. and Chen, S.X., 2012, Two sample test for high dimensional covariance matrices, Anal. Stat. 40: 908-940.
 Bai, Z., Jiang, D., Yao, J.F. and Zheng, S., 2009, Corrections to LRT on large-dimensional covariance matrices by RMT, Ann. Statist. 37: 3822-3840.
 Srivastava, M.S., Yanagihara, H. and Kunokawa, T., 2014, Test for covariance matrices in high dimension with less sample size, J. Multivariate Anal. 130: 289-309.
 Chaipitak, S., 2013, Tests for covariance Matrices with High-Dimensional Data, Ph.D. Thesis, National Institute of Development Administration, Bangkok, 124 p.
 Chochran, W.G., 1954, Some methods for strengthening the common test, Biometrics 10: 417-451.
 Chen, S.X. and Qin, Y.L., 2010, A two-sample test for high-dimensional data with applications to gene-set testing, Anal. Stat. 38: 808-835.
 Cai, T., Lui, W. and Xai, Y., 2013 Two-sample covariance matrix testing and support recovery in high-dimensional and sparse settings, J. Amer. Stat. Assoc. 108: 265-277.