Estimation of intra-cluster correlation coefficient via the failure of Bartlett's second identity

Tsung Shan Tsou; Wan Chen Chen

doi:10.1007/s00180-012-0372-7

Estimation of intra-cluster correlation coefficient via the failure of Bartlett's second identity

Tsung Shan Tsou, Wan Chen Chen

Graduate Institute of Statistics

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

Abstract

A new means of estimating the correlation coefficient for cluster binary data in the regression settings is introduced. The creation of this method is founded upon the violation of Bartlett's second identity when adopting the binomial distributions to model binary data that are correlated. The new methodology applies to any sensible link functions that connect the success probability and covariates. One can easily implement the procedure by using any statistical software providing the naïve and the sandwich covariance matrices for regression parameter estimates. Simulations and real data analyses are used to demonstrate the efficacy of our new procedure.

Original language	English
Pages (from-to)	1681-1698
Number of pages	18
Journal	Computational Statistics
Volume	28
Issue number	4
DOIs	https://doi.org/10.1007/s00180-012-0372-7
State	Published - Aug 2013

Keywords

Bartlett's second identity
Correlated binary data
Correlation coefficient
Fisher information matrix

Access to Document

10.1007/s00180-012-0372-7

Cite this

@article{4d2d149a7a564a309baf1a1d1b87b34a,

title = "Estimation of intra-cluster correlation coefficient via the failure of Bartlett's second identity",

abstract = "A new means of estimating the correlation coefficient for cluster binary data in the regression settings is introduced. The creation of this method is founded upon the violation of Bartlett's second identity when adopting the binomial distributions to model binary data that are correlated. The new methodology applies to any sensible link functions that connect the success probability and covariates. One can easily implement the procedure by using any statistical software providing the na{\"i}ve and the sandwich covariance matrices for regression parameter estimates. Simulations and real data analyses are used to demonstrate the efficacy of our new procedure.",

keywords = "Bartlett's second identity, Correlated binary data, Correlation coefficient, Fisher information matrix",

author = "Tsou, {Tsung Shan} and Chen, {Wan Chen}",

note = "Funding Information: This work is supported by grant NSC 100-2118-M-008-001-MY2 of the National Science Council, and National Central University-Land Seed Hospital Joint Research grant NCU-LSH-100-A-005, Taiwan, R.O.C.",

year = "2013",

month = aug,

doi = "10.1007/s00180-012-0372-7",

language = "???core.languages.en_GB???",

volume = "28",

pages = "1681--1698",

journal = "Computational Statistics",

issn = "0943-4062",

number = "4",

}

TY - JOUR

T1 - Estimation of intra-cluster correlation coefficient via the failure of Bartlett's second identity

AU - Tsou, Tsung Shan

AU - Chen, Wan Chen

N1 - Funding Information: This work is supported by grant NSC 100-2118-M-008-001-MY2 of the National Science Council, and National Central University-Land Seed Hospital Joint Research grant NCU-LSH-100-A-005, Taiwan, R.O.C.

PY - 2013/8

Y1 - 2013/8

N2 - A new means of estimating the correlation coefficient for cluster binary data in the regression settings is introduced. The creation of this method is founded upon the violation of Bartlett's second identity when adopting the binomial distributions to model binary data that are correlated. The new methodology applies to any sensible link functions that connect the success probability and covariates. One can easily implement the procedure by using any statistical software providing the naïve and the sandwich covariance matrices for regression parameter estimates. Simulations and real data analyses are used to demonstrate the efficacy of our new procedure.

AB - A new means of estimating the correlation coefficient for cluster binary data in the regression settings is introduced. The creation of this method is founded upon the violation of Bartlett's second identity when adopting the binomial distributions to model binary data that are correlated. The new methodology applies to any sensible link functions that connect the success probability and covariates. One can easily implement the procedure by using any statistical software providing the naïve and the sandwich covariance matrices for regression parameter estimates. Simulations and real data analyses are used to demonstrate the efficacy of our new procedure.

KW - Bartlett's second identity

KW - Correlated binary data

KW - Correlation coefficient

KW - Fisher information matrix

UR - http://www.scopus.com/inward/record.url?scp=84881062094&partnerID=8YFLogxK

U2 - 10.1007/s00180-012-0372-7

DO - 10.1007/s00180-012-0372-7

M3 - 期刊論文

AN - SCOPUS:84881062094

SN - 0943-4062

VL - 28

SP - 1681

EP - 1698

JO - Computational Statistics

JF - Computational Statistics

IS - 4

ER -

Estimation of intra-cluster correlation coefficient via the failure of Bartlett's second identity

Abstract

Keywords

Access to Document

Fingerprint

Cite this