Diagnostics for binomial response models using power divergence statistics

Edward J. Bedrick; Chih Ling Tsai

doi:10.1016/0167-9473(93)90171-O

Diagnostics for binomial response models using power divergence statistics

Edward J. Bedrick, Chih Ling Tsai

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

2 Scopus citations

Abstract

The family of power divergence statistics (Cressie and Read, 1984) is used to construct diagnostics for dose-response models. Single case deletion diagnostics based on maximum likelihood, minimum chi-square, and minimum Hellinger distance estimates are included in this class. One-step approximations to outlier tests and influence measures are given. We also discuss method to evaluate the sensitivity of power divergence estimates and diagnostics to the choice of the power divergence parameter. The techniques are illustrated on a data set.

Original language	English (US)
Pages (from-to)	381-392
Number of pages	12
Journal	Computational Statistics and Data Analysis
Volume	15
Issue number	4
DOIs	https://doi.org/10.1016/0167-9473(93)90171-O
State	Published - May 1993
Externally published	Yes

Keywords

Cook's distance
Logistic regression
Minimum Hellinger distance estimation
Minimum chi-square estimation
Outlier
Residual

ASJC Scopus subject areas

Statistics and Probability
Computational Mathematics
Computational Theory and Mathematics
Applied Mathematics

Access to Document

10.1016/0167-9473(93)90171-O

Cite this

@article{d3ce05189238499aa255f6237dd85e7f,

title = "Diagnostics for binomial response models using power divergence statistics",

abstract = "The family of power divergence statistics (Cressie and Read, 1984) is used to construct diagnostics for dose-response models. Single case deletion diagnostics based on maximum likelihood, minimum chi-square, and minimum Hellinger distance estimates are included in this class. One-step approximations to outlier tests and influence measures are given. We also discuss method to evaluate the sensitivity of power divergence estimates and diagnostics to the choice of the power divergence parameter. The techniques are illustrated on a data set.",

keywords = "Cook's distance, Logistic regression, Minimum Hellinger distance estimation, Minimum chi-square estimation, Outlier, Residual",

author = "Bedrick, {Edward J.} and Tsai, {Chih Ling}",

year = "1993",

month = may,

doi = "10.1016/0167-9473(93)90171-O",

language = "English (US)",

volume = "15",

pages = "381--392",

journal = "Computational Statistics and Data Analysis",

issn = "0167-9473",

publisher = "Elsevier B.V.",

number = "4",

}

TY - JOUR

T1 - Diagnostics for binomial response models using power divergence statistics

AU - Bedrick, Edward J.

AU - Tsai, Chih Ling

PY - 1993/5

Y1 - 1993/5

N2 - The family of power divergence statistics (Cressie and Read, 1984) is used to construct diagnostics for dose-response models. Single case deletion diagnostics based on maximum likelihood, minimum chi-square, and minimum Hellinger distance estimates are included in this class. One-step approximations to outlier tests and influence measures are given. We also discuss method to evaluate the sensitivity of power divergence estimates and diagnostics to the choice of the power divergence parameter. The techniques are illustrated on a data set.

AB - The family of power divergence statistics (Cressie and Read, 1984) is used to construct diagnostics for dose-response models. Single case deletion diagnostics based on maximum likelihood, minimum chi-square, and minimum Hellinger distance estimates are included in this class. One-step approximations to outlier tests and influence measures are given. We also discuss method to evaluate the sensitivity of power divergence estimates and diagnostics to the choice of the power divergence parameter. The techniques are illustrated on a data set.

KW - Cook's distance

KW - Logistic regression

KW - Minimum Hellinger distance estimation

KW - Minimum chi-square estimation

KW - Outlier

KW - Residual

UR - http://www.scopus.com/inward/record.url?scp=43949169886&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=43949169886&partnerID=8YFLogxK

U2 - 10.1016/0167-9473(93)90171-O

DO - 10.1016/0167-9473(93)90171-O

M3 - Article

AN - SCOPUS:43949169886

SN - 0167-9473

VL - 15

SP - 381

EP - 392

JO - Computational Statistics and Data Analysis

JF - Computational Statistics and Data Analysis

IS - 4

ER -

Diagnostics for binomial response models using power divergence statistics

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this