Zusammenfassung
Dieser Artikel liefert einen Literaturüberblick für zwei robuste statistische Methoden — "trimming" und "Winsorization", welche zuerst vorgeschlagen wurden, um Lageparameter zu schätzen, später aber auf andere Schätz- und Testprobleme erweitert wurden. Die Anwendbarkeit der Methoden für normale und "long-tailed"-Verteilungen wird diskutiert.
Summary
This paper provides a literature review for robust statistical procedures — trimming and Winsorization — that were first proposed for estimating location, but were later extended to other estimation and testing problems. Performance of these techniques under normal and long-tailed distributions are discussed.
Résumé
Cet article fournit un revoir de la litterature pour deux procedures statistiques — "trimming" et Winsorization" — qui etaient, d’abord, proposees a estimer position, mais qui etaient plus tard etendues a d’autres problemes d’estimation et de verification. On discute le fonctionnement de ces techniques avec les distributions "normale" et "long-tailed".
Резюме
В Этой статье дан литературный обзор для двух статистических метод: «тримминга» и «винзоризейщн». Сначала они были выработаны только для оценки параметра положения а потом расщирились на другие проблемы оценки и проверки. Обсуждается применимость Этих метод для нормального распределения и распределения «лёнг-тайльд.
References
Andrews, D.F., P.J. Bickel, F.R. Hample, P.J. Huber, W.H. Rogers, and J.W. Tukey (1972).Robust estimates of location: Survey and Advances. Princeton Univ. Press.
Bickel, P.J. (1965). On some robust estimate of location.Ann. Math. Statist., 36, 847–858.
Chen, E.H. (1969). Winsorization and trimming techniques applied to linear regression analysis. Ph.D. dissertation, Univ. of Calif., Los Angeles.
Chen, E.H. and W.J. Dixon (1972). Mean and MSE for the Winsorized mean corresponding to the shortest 95 % confidence interval. Unpublished manuscript.
Chen, E.H. and W.J. Dixon (1972). Estimates of parameters of a censored regression sample.J. Amer. Statist. Assoc., 67, 664–671.
Dixon, W.J. (1957). Estimates of the mean and standard deviation of a normal population.Ann. Math. Statist., 28, 806–809.
Dixon, W.J. (1960). Simplified estimation from censored normal samples.Ann. Math. Statist., 31, 385–391.
Dixon, W.J. and J.W. Tukey (1969). Approximate behavior of the distribution of Winsorized t. (Trimming/Winsorization 2)Technometrics, 10, 83–98.
Dixon, W.J. and F.J. Massey, Jr. (1969).Introduction to statistical analysis, 3rd ed., New York: McGraw-Hill.
Fridshal, D. and H.O. Posten (1966). Bibliography on statistical robustness and related topics. Research report No. 16, Dept. of Statistics, Univ. of Connecticut, Storrs, Conn.
Grovindarajulu, J. and R.T. Leslie (1972). Annotated bibliography on robustness studies. Vital and health statistics, data evaluation and method research, series 2, No. 51.
Harris, T.E. and J.W. Tukey (1949). Development of large sample measures of location and scale which are relatively insensitive to contamination. Memorandum Report 31, Statistical research group, Princeton Univ.
Hatch, L.O. and H.O. Posten (1966). Robustness of Student-procedure: A survey. Research report No. 24, Dept. of Statistics, Univ. of Connecticut.
Huber, P.J. (1972). Robust statistics: A review.Ann. Math. Statist., 43, 1041–1067.
Hyrenius, H. and I. Adolfson, et al. (1964). Selected bibliography on nonnormality. Publication No. 12, Dept. of Statistics, Univ. of Gothenburg, Sweden.
Jaeckel, L.A. (1971). Some flexible estimates of location.Ann. Math. Statist., 42, 1540–1552.
Johnson, N.L. (1949). Systems of frequency curves generated by methods of translation.Biometrika, 36, 149–176.
McLaughlin, D.H. and J.W. Tukey (1961). The variance of symmetrically trimmed samples from normal populations, and its estimation from such trimmed samples. (Trimming/Winsorization I), Princeton Univ. Technical Report No. 42.
Stigler, S.M. (1972). The asymptotic distribution of the trimmed mean. Technical report No. 294, Dept. of Statistics, The Univ. of Wisconsin, Madison, Wisconsin.
Tukey, J.W. (1962). The future of data analysis.Ann. Math. Statist., 33, 1–67.
Tukey, J.W. and D.H. McLaughlin (1963). Less vulnerable confidence and significance procedures for location based on a single sample: Trimming/Winsorization I.Sankhyā, A, 25, 331–352.
Walsh, J.E. (1949). Some significance tests for the median which are valid under very general conditions.Ann. Math. Statist., 20, 64–81.
Wang, Y.Y. (1971). Probabilities of type I errors of the Welch tests for the Behrens-Fisher problem.J. Amer. Statist. Assoc., 66, 605–608.
Welch, B.L. (1936). Specification of rules for rejecting two variable a product, with particular reference to an electric lamp problem.J. Roy. Statist. Soc., Supp. 3, 29–48.
Welch, B.L. (1938). The significance of the difference between two means when the population variances are unequal.Biometrika, 29, 350–62.
Welch, B.L. (1949). In Appendix of Tables for use in comparisons whose accuracy involves two variances separately estimated by A.A. Aspin.Biometrika, 36, 290–96.
Wonnacott, T.H. (1963). A Monte-Carlo method of obtaining the power of certain tests of location. Ph.D.dissertation, Princeton Univ.
Yale, Coralee (1970). An application of Winsorization to linear regression analysis. Masters thesis, Univ. of. Calif., Los Angeles.
Yuen, K.K. (1971). A note on Winsorized t.J. Roy. Statist. Soc., C. 20, 297–304.
Yuen, K.K. (1972). The two-sample trimmed t. Ph.D. dissertation, Univ. of Calif., Los Angeles.
Yuen, K.K. (1972). Power comparisons of single-sample Winsorized, trimmed, and Student’s t. Submitted for publication.
Yuen, K.K. and W.J. Dixon (1973). The approximate behavior and performance of the two-sample trimmed t.Biometrika, 60, 2, 369–374.
Yuen, K.K. (1974). The two-sample trimmed t for unequal population variances.Biometrika, 61, 165–170.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Dixon, W.J., Yuen, K.K. Trimming and winsorization: A review. Statistische Hefte 15, 157–170 (1974). https://doi.org/10.1007/BF02922904
Issue Date:
DOI: https://doi.org/10.1007/BF02922904