×
Paul, S. R.

On the detection of unusual points in regression. (English) Zbl 0607.62033

Biom. J. 29, 147-152 (1987).
Some procedures, based on studentized residuals and the diagonal elements of the hat matrix, for the detection of unusual points in regression are discussed. These procedures are quite simple but effective to uncover unusual structure of the data. Some examples are given.

MSC:

62F35 Robustness and adaptive procedures (parametric inference)
62J05 Linear regression; mixed models

Keywords:

single outlier test; single high leverage point test; studentized residuals; diagonal elements; hat matrix; detection of unusual points in regression

Software:

IMSL Numerical Libraries
Cite Review PDF
Full Text: DOI

References:

[1] Atkinson, Biometrika 68 pp 13– (1981)
[2] Andrews, J. Roy. Statist. Soc. B40 pp 85– (1978)
[3] and , 1980: Regression Diagnostics: Identifying Influential Data and Source of Collinearity. Wiley, New York. · Zbl 0479.62056 · doi:10.1002/0471725153
[4] 1965: Statistical Theory and Methodology in Science and Engineering, 2nd Edn. Wiley, New York.
[5] Cook, Technometrics 22 pp 495– (1980)
[6] 1982: Graphical displays for the diagnosis of influential observations and outliers in linear regressio n. Division of Biostatistics, University of Rochester, School of Medicine and Dentistry, Rochester, New York, 14642 (unpublished report).
[7] Dempster, Ann. Statist. 9 pp 945– (1981)
[8] Draper, Technometrics 23 pp 21– (1981)
[9] Gentleman, Biometrics 31 pp 387– (1975)
[10] Hoaglin, Amer. Statist. 32 pp 17– (1978)
[11] 1981: Robust Statistics. New York: John Wiley. · Zbl 0536.62025 · doi:10.1002/0471725250
[12] International Mathematical and Statistical Libraries (1979), IMSL Library Reference Manual (Vol. 2, 7th ed.).
[13] Pauil, The Statistician 32 pp 417– (1983)
[14] Wood, Technometrics 15 pp 677– (1973)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.