Approximate test for testing a null variance ratio in the Unbalanced One-Way Random Model

Abstract

The approximate test for testing the signicance of the random effect is presented in the unbalanced one-way random model in which both random effects and errors are from non-normal universes. The test is based on the asymptotic distribution of the F-statistic. Under the condition that the number of groups tends to infinity while the average of powers of the group sizes is bounded, the asymptotic distribution of the F-statistic is obtained. Robustness of the proposed test is given.

Keywords

• Asymptotic normality,F-test,robustness of the approximate F-test,rondom effects

References

1. Akritas, M and Arnold, S., Asymptotic for analysis of variance when the number of levels is large, J. Amer Statist Assoc. 95, (2000), 212-226.
2. Akritas, M.G. and Papdatos, N., Heteroscedastic one-way ANOVA and lack-of-fit tests, J. Amer Statist. Assoc., 99, (2004), 368-382.
3. Güven, B., Testing for random effect in the Fuller-Battese model, Statistics, 48, (2014), 802-814.
4. Rudin, W., Principles of Mathematical Analysis, New York, McGraw-Hill, 1976.
5. Searle, S. R., Linear Models for Unbalanced Data, New York, Wiley, 1988.
6. Scheffe, H., The Analysis of Variance, New York Wiley, 1959
7. Westfall, P., A comparison of variance component estimates for arbitrary underlying distributions, J. Amer. Statist. Assoc., 62, (1987), 866-874.
8. Westfall, P., Robustness and power of tests for a null variance ratio, Biometrika, 75; (1988), 207-214.

9. van der Vaart, A.W., Asymptotic statistics, New York, Cambridge 1998.