Abstract
In this paper, we given an approach for detecting one or more outliers in randomized linear model. The likelihood ratio test statistic and its distributions under the null hypothesis and the alternative hypothesis are given. Furthermore, the robustness of the test statistic in a certain sense is proved. Finally, the optimality properties of the test are derived.
Similar content being viewed by others
References
Srikaman, K.S., Testing for the single outlier in a regression model,Snakhyā,A 23 (1961), 251–260.
Barnett, V. and Lewis, T., Outlier Statistical Data. Wiley, New York, 1978.
Butler, R., Outlier discordancy tests in the normal linear model,J. R. Statist. Soc. B 45:1 (1983), 120–132.
Hartley, H. O. and Rao, C. R., Maximum-likelihood estimation for the mixed analysis of variance model,Biometrika,54 (1967), 93–108.
Shi Lei and Chen Jianbao, Influence analysis in stochatistic linear model, Technical Report of The 4th China-Japan Symposium on Statistics, 1991.
Wang Songgui, The Theory and Its Application of Linear Model, Anhui Education Publishing House, 1987.
Chen Xiru, The Introduction of Mathematical Statistics, Science Publishing House, 1981.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Liming, X., Lei, S. Outlier test in randomized linear model. Appl. Math. 9, 65–75 (1994). https://doi.org/10.1007/BF02662027
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02662027
Key words
- Randomized Linear Model
- Outliers
- Likelihood Ratio Test
- Uniformly Most Powerfull Invariant (UMPI)
- Uniformly Most Powerfull Unbiased (UMPU)