Assessing influence on predictions from generalized linear models

William Thomas; R. Dennis Cook

doi:10.1080/00401706.1990.10484593

Assessing influence on predictions from generalized linear models

William Thomas, R. Dennis Cook

Statistics (Twin Cities)

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

81 Scopus citations

Abstract

Influence diagnostics for predictions from a normal linear model examine the effect of deleting a single case on either the point prediction or the predictive density function. Instead of deleting cases, we apply the local influence method of Cook (1986) to assess the effect of small perturbations of continuous data on a specified point prediction from a generalized linear model. Based on local perturbations of the vector of responses, case weights, explanatory variables, or the components of one case, the diagnostics can detect different kinds of influence. Some of the diagnostics are illustrated with an example and compared to standard diagnostic methods.

Original language	English (US)
Pages (from-to)	59-65
Number of pages	7
Journal	Technometrics
Volume	32
Issue number	1
DOIs	https://doi.org/10.1080/00401706.1990.10484593
State	Published - Feb 1990

Bibliographical note

Funding Information:
Much of this article is taken from Thomas’s Ph.D. thesis, written under the supervision of Cook at the University of Minnesota. This work was supported in part by grants from the National Science Foundation.

Keywords

Data perturbation
Diagnostics
Leverage
Local influence

Publisher link

10.1080/00401706.1990.10484593

Find It

Find It at U of M Libraries

Cite this

@article{2ca441408a534f36bc6430c82df99661,

title = "Assessing influence on predictions from generalized linear models",

abstract = "Influence diagnostics for predictions from a normal linear model examine the effect of deleting a single case on either the point prediction or the predictive density function. Instead of deleting cases, we apply the local influence method of Cook (1986) to assess the effect of small perturbations of continuous data on a specified point prediction from a generalized linear model. Based on local perturbations of the vector of responses, case weights, explanatory variables, or the components of one case, the diagnostics can detect different kinds of influence. Some of the diagnostics are illustrated with an example and compared to standard diagnostic methods.",

keywords = "Data perturbation, Diagnostics, Leverage, Local influence",

author = "William Thomas and Cook, {R. Dennis}",

note = "Funding Information: Much of this article is taken from Thomas{\textquoteright}s Ph.D. thesis, written under the supervision of Cook at the University of Minnesota. This work was supported in part by grants from the National Science Foundation.",

year = "1990",

month = feb,

doi = "10.1080/00401706.1990.10484593",

language = "English (US)",

volume = "32",

pages = "59--65",

journal = "Technometrics",

issn = "0040-1706",

publisher = "American Statistical Association",

number = "1",

}

TY - JOUR

T1 - Assessing influence on predictions from generalized linear models

AU - Thomas, William

AU - Cook, R. Dennis

N1 - Funding Information: Much of this article is taken from Thomas’s Ph.D. thesis, written under the supervision of Cook at the University of Minnesota. This work was supported in part by grants from the National Science Foundation.

PY - 1990/2

Y1 - 1990/2

N2 - Influence diagnostics for predictions from a normal linear model examine the effect of deleting a single case on either the point prediction or the predictive density function. Instead of deleting cases, we apply the local influence method of Cook (1986) to assess the effect of small perturbations of continuous data on a specified point prediction from a generalized linear model. Based on local perturbations of the vector of responses, case weights, explanatory variables, or the components of one case, the diagnostics can detect different kinds of influence. Some of the diagnostics are illustrated with an example and compared to standard diagnostic methods.

AB - Influence diagnostics for predictions from a normal linear model examine the effect of deleting a single case on either the point prediction or the predictive density function. Instead of deleting cases, we apply the local influence method of Cook (1986) to assess the effect of small perturbations of continuous data on a specified point prediction from a generalized linear model. Based on local perturbations of the vector of responses, case weights, explanatory variables, or the components of one case, the diagnostics can detect different kinds of influence. Some of the diagnostics are illustrated with an example and compared to standard diagnostic methods.

KW - Data perturbation

KW - Diagnostics

KW - Leverage

KW - Local influence

UR - http://www.scopus.com/inward/record.url?scp=0025388464&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0025388464&partnerID=8YFLogxK

U2 - 10.1080/00401706.1990.10484593

DO - 10.1080/00401706.1990.10484593

M3 - Article

AN - SCOPUS:0025388464

SN - 0040-1706

VL - 32

SP - 59

EP - 65

JO - Technometrics

JF - Technometrics

IS - 1

ER -

Assessing influence on predictions from generalized linear models

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this