Model selection criteria for loglinear models

E. J. Bedrick; W. K. Crandall

doi:10.1111/j.1467-842X.2010.00593.x

Model selection criteria for loglinear models

E. J. Bedrick, W. K. Crandall

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

1 Scopus citations

Abstract

Considerable work has been devoted to developing model selection criteria for normal theory regression models. Less attention has been paid to discrete data. We develop two loglinear model selection criteria for Poisson counts. These criteria are based on an estimated bias adjustment of the Akaike information criterion. We observe in a simulation study that the corrected statistics provide good model choices and relatively accurate estimates of the mean structure.

Original language	English (US)
Pages (from-to)	439-449
Number of pages	11
Journal	Australian and New Zealand Journal of Statistics
Volume	52
Issue number	4
DOIs	https://doi.org/10.1111/j.1467-842X.2010.00593.x
State	Published - Dec 2010
Externally published	Yes

Keywords

AIC
BIC
Contingency table
Multinomial model
Poisson counts

ASJC Scopus subject areas

Statistics and Probability
Statistics, Probability and Uncertainty

Access to Document

10.1111/j.1467-842X.2010.00593.x

Cite this

@article{ea691caf76a6488995fdda965274930e,

title = "Model selection criteria for loglinear models",

abstract = "Considerable work has been devoted to developing model selection criteria for normal theory regression models. Less attention has been paid to discrete data. We develop two loglinear model selection criteria for Poisson counts. These criteria are based on an estimated bias adjustment of the Akaike information criterion. We observe in a simulation study that the corrected statistics provide good model choices and relatively accurate estimates of the mean structure.",

keywords = "AIC, BIC, Contingency table, Multinomial model, Poisson counts",

author = "Bedrick, {E. J.} and Crandall, {W. K.}",

year = "2010",

month = dec,

doi = "10.1111/j.1467-842X.2010.00593.x",

language = "English (US)",

volume = "52",

pages = "439--449",

journal = "Australian and New Zealand Journal of Statistics",

issn = "1369-1473",

number = "4",

}

TY - JOUR

T1 - Model selection criteria for loglinear models

AU - Bedrick, E. J.

AU - Crandall, W. K.

PY - 2010/12

Y1 - 2010/12

N2 - Considerable work has been devoted to developing model selection criteria for normal theory regression models. Less attention has been paid to discrete data. We develop two loglinear model selection criteria for Poisson counts. These criteria are based on an estimated bias adjustment of the Akaike information criterion. We observe in a simulation study that the corrected statistics provide good model choices and relatively accurate estimates of the mean structure.

AB - Considerable work has been devoted to developing model selection criteria for normal theory regression models. Less attention has been paid to discrete data. We develop two loglinear model selection criteria for Poisson counts. These criteria are based on an estimated bias adjustment of the Akaike information criterion. We observe in a simulation study that the corrected statistics provide good model choices and relatively accurate estimates of the mean structure.

KW - AIC

KW - BIC

KW - Contingency table

KW - Multinomial model

KW - Poisson counts

UR - http://www.scopus.com/inward/record.url?scp=78650642800&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78650642800&partnerID=8YFLogxK

U2 - 10.1111/j.1467-842X.2010.00593.x

DO - 10.1111/j.1467-842X.2010.00593.x

M3 - Article

AN - SCOPUS:78650642800

SN - 1369-1473

VL - 52

SP - 439

EP - 449

JO - Australian and New Zealand Journal of Statistics

JF - Australian and New Zealand Journal of Statistics

IS - 4

ER -

Model selection criteria for loglinear models

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this