Estimation of variation after outlier rejection

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Outliers are common in many sources of data used for estimation of variation; for example, they are a constant problem in data used to determine the precision of analytical methods. Ordinary analysis of unaltered data is often undesirable because the value of an estimate of variation can be essentially determined by only a few points. This paper shows that outlier rejection followed by an ordinary analysis of the remaining data is not acceptable method either - when any data are actually rejected, the variance of the remaining points is badly biased down, even when the distribution from which the data are generated is outlier-prone. By comparison, the robust method of Rocke (1983) performs well. Maximum likelihood for mixtures of normals is promising, but requires further development before its use is practical in small samples.

Original languageEnglish (US)
Pages (from-to)9-20
Number of pages12
JournalComputational Statistics and Data Analysis
Volume13
Issue number1
StatePublished - Jan 1992

Fingerprint

Rejection
Maximum likelihood
Outlier
Robust Methods
Analytical Methods
Small Sample
Biased
Maximum Likelihood
Outliers
Estimate

Keywords

  • Components of variance
  • Maximum likelihood estimation
  • Mixture of normals

ASJC Scopus subject areas

  • Statistics and Probability
  • Numerical Analysis
  • Computational Mathematics
  • Electrical and Electronic Engineering
  • Statistics, Probability and Uncertainty
  • Computational Theory and Mathematics

Cite this

Estimation of variation after outlier rejection. / Rocke, David M.

In: Computational Statistics and Data Analysis, Vol. 13, No. 1, 01.1992, p. 9-20.

Research output: Contribution to journalArticle

@article{7a4018a675814c01baaa179730a9edce,
title = "Estimation of variation after outlier rejection",
abstract = "Outliers are common in many sources of data used for estimation of variation; for example, they are a constant problem in data used to determine the precision of analytical methods. Ordinary analysis of unaltered data is often undesirable because the value of an estimate of variation can be essentially determined by only a few points. This paper shows that outlier rejection followed by an ordinary analysis of the remaining data is not acceptable method either - when any data are actually rejected, the variance of the remaining points is badly biased down, even when the distribution from which the data are generated is outlier-prone. By comparison, the robust method of Rocke (1983) performs well. Maximum likelihood for mixtures of normals is promising, but requires further development before its use is practical in small samples.",
keywords = "Components of variance, Maximum likelihood estimation, Mixture of normals",
author = "Rocke, {David M}",
year = "1992",
month = "1",
language = "English (US)",
volume = "13",
pages = "9--20",
journal = "Computational Statistics and Data Analysis",
issn = "0167-9473",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - Estimation of variation after outlier rejection

AU - Rocke, David M

PY - 1992/1

Y1 - 1992/1

N2 - Outliers are common in many sources of data used for estimation of variation; for example, they are a constant problem in data used to determine the precision of analytical methods. Ordinary analysis of unaltered data is often undesirable because the value of an estimate of variation can be essentially determined by only a few points. This paper shows that outlier rejection followed by an ordinary analysis of the remaining data is not acceptable method either - when any data are actually rejected, the variance of the remaining points is badly biased down, even when the distribution from which the data are generated is outlier-prone. By comparison, the robust method of Rocke (1983) performs well. Maximum likelihood for mixtures of normals is promising, but requires further development before its use is practical in small samples.

AB - Outliers are common in many sources of data used for estimation of variation; for example, they are a constant problem in data used to determine the precision of analytical methods. Ordinary analysis of unaltered data is often undesirable because the value of an estimate of variation can be essentially determined by only a few points. This paper shows that outlier rejection followed by an ordinary analysis of the remaining data is not acceptable method either - when any data are actually rejected, the variance of the remaining points is badly biased down, even when the distribution from which the data are generated is outlier-prone. By comparison, the robust method of Rocke (1983) performs well. Maximum likelihood for mixtures of normals is promising, but requires further development before its use is practical in small samples.

KW - Components of variance

KW - Maximum likelihood estimation

KW - Mixture of normals

UR - http://www.scopus.com/inward/record.url?scp=38249014752&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38249014752&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:38249014752

VL - 13

SP - 9

EP - 20

JO - Computational Statistics and Data Analysis

JF - Computational Statistics and Data Analysis

SN - 0167-9473

IS - 1

ER -