(Received July 22, 1991; revised July 27, 1992)
Abstract. In this paper we address the problem of testing the equality of k intraclass correlation coefficients based on samples from independent p-variate normal populations, and explore various aspects of optimality through invariance. A UMPIU test is derived for k = 2, and LMMPIU test of SenGupta and Vermeire (1986) is indicated for k > 2. Several approximately optimum invariant tests are also proposed. The tests are compared with the approximate LR tests and Fisher's Z-tests derived in Konishi and Gupta (1987, 1989). As expected, the performance of the proposed tests turns out to be quite satisfactory and superior to the LR tests and Z-tests.
Key words and phrases: Intraclass correlation, invariance, locally most powerful invariant unbiased test, uniformly most powerful invariant unbiased test.
Source ( TeX , DVI , PS )