Outlier tests for logistic regression: A conditional approach

Edward J. Bedrick; Joe R. Hill

doi:10.1093/biomet/77.4.815

Outlier tests for logistic regression: A conditional approach

Edward J. Bedrick, Joe R. Hill

Research output: Contribution to journal › Article › peer-review

16 Scopus citations

Abstract

We consider exact conditional methods for identifying outliers in logistic regression data. Tests for a single outlier and multiple outliers are developed assuming a logistic slippage model. The p-values for these tests are determined using an explicit enumeration of all possible responses consistent with the observed value of the sufficient statistic. Justifications are given for preferring this computationally intensive approach to standard methods based on asymptotic approximations. The techniques are applied to two examples.

Original language	English (US)
Pages (from-to)	815-827
Number of pages	13
Journal	Biometrika
Volume	77
Issue number	4
DOIs	https://doi.org/10.1093/biomet/77.4.815
State	Published - Dec 1990
Externally published	Yes

Keywords

Exact inference
P-value
Residual

ASJC Scopus subject areas

Statistics and Probability
General Mathematics
Agricultural and Biological Sciences (miscellaneous)
General Agricultural and Biological Sciences
Statistics, Probability and Uncertainty
Applied Mathematics

Access to Document

10.1093/biomet/77.4.815

Cite this

@article{d760f50d12304d5ead3e46e755b95443,

title = "Outlier tests for logistic regression: A conditional approach",

abstract = "We consider exact conditional methods for identifying outliers in logistic regression data. Tests for a single outlier and multiple outliers are developed assuming a logistic slippage model. The p-values for these tests are determined using an explicit enumeration of all possible responses consistent with the observed value of the sufficient statistic. Justifications are given for preferring this computationally intensive approach to standard methods based on asymptotic approximations. The techniques are applied to two examples.",

keywords = "Exact inference, P-value, Residual",

author = "Bedrick, {Edward J.} and Hill, {Joe R.}",

year = "1990",

month = dec,

doi = "10.1093/biomet/77.4.815",

language = "English (US)",

volume = "77",

pages = "815--827",

journal = "Biometrika",

issn = "0006-3444",

publisher = "Oxford University Press",

number = "4",

}

TY - JOUR

T1 - Outlier tests for logistic regression

T2 - A conditional approach

AU - Bedrick, Edward J.

AU - Hill, Joe R.

PY - 1990/12

Y1 - 1990/12

N2 - We consider exact conditional methods for identifying outliers in logistic regression data. Tests for a single outlier and multiple outliers are developed assuming a logistic slippage model. The p-values for these tests are determined using an explicit enumeration of all possible responses consistent with the observed value of the sufficient statistic. Justifications are given for preferring this computationally intensive approach to standard methods based on asymptotic approximations. The techniques are applied to two examples.

AB - We consider exact conditional methods for identifying outliers in logistic regression data. Tests for a single outlier and multiple outliers are developed assuming a logistic slippage model. The p-values for these tests are determined using an explicit enumeration of all possible responses consistent with the observed value of the sufficient statistic. Justifications are given for preferring this computationally intensive approach to standard methods based on asymptotic approximations. The techniques are applied to two examples.

KW - Exact inference

KW - P-value

KW - Residual

UR - http://www.scopus.com/inward/record.url?scp=0040124737&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0040124737&partnerID=8YFLogxK

U2 - 10.1093/biomet/77.4.815

DO - 10.1093/biomet/77.4.815

M3 - Article

AN - SCOPUS:0040124737

SN - 0006-3444

VL - 77

SP - 815

EP - 827

JO - Biometrika

JF - Biometrika

IS - 4

ER -

Outlier tests for logistic regression: A conditional approach

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this