Abstract

This study proposed a test for the equality of two covariance matrices from two independent multivariate normal populations with high-dimensional data. The test statistic is based on unbiased and consistent estimator of the ratio between the sums of squares of covariance matrix elements. Under the null hypothesis, the proposed test statistic is asymptotically standard normal distributed when the number of variables and the sample sizes go together to infinity. Simulation study is conducted to investigate the performance of the proposed test statistic. The results showed that the proposed test is superior to the other three tests appeared in the literature for various patterns of common covariance matrix. Finally, two real data sets are analyzed to illustrate the application of our theoretical results.