Estimates for cell counts and common odds ratio in three-way contingency tables by homogeneous log-linear models with missing data

Haresh D. Rochani, Robert L. Vogel, Hani M. Samawi, Daniel F Linder

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Missing observations often occur in cross-classified data collected during observational, clinical, and public health studies. Inappropriate treatment of missing data can reduce statistical power and give biased results. This work extends the Baker, Rosenberger and Dersimonian modeling approach to compute maximum likelihood estimates for cell counts in three-way tables with missing data, and studies the association between two dichotomous variables while controlling for a third variable in 2 × 2 × K tables. This approach is applied to the Behavioral Risk Factor Surveillance System data. Simulation studies are used to investigate the efficiency of estimation of the common odds ratio.

Original languageEnglish (US)
Pages (from-to)51-65
Number of pages15
JournalAStA Advances in Statistical Analysis
Volume101
Issue number1
DOIs
StatePublished - Jan 1 2017

Fingerprint

Common ratio
Log-linear Models
Odds Ratio
Contingency Table
Public health
linear model
Missing Data
contingency
Maximum likelihood
Tables
Count
Missing Observations
Statistical Power
Cell
Public Health
Risk Factors
Maximum Likelihood Estimate
Estimate
Surveillance
Biased

Keywords

  • Common odds ratio
  • Contingency table
  • Cross-classified data
  • Log-linear model
  • Maximum likelihood method
  • Missing data
  • Three-way table

ASJC Scopus subject areas

  • Analysis
  • Statistics and Probability
  • Modeling and Simulation
  • Social Sciences (miscellaneous)
  • Economics and Econometrics
  • Applied Mathematics

Cite this

Estimates for cell counts and common odds ratio in three-way contingency tables by homogeneous log-linear models with missing data. / Rochani, Haresh D.; Vogel, Robert L.; Samawi, Hani M.; Linder, Daniel F.

In: AStA Advances in Statistical Analysis, Vol. 101, No. 1, 01.01.2017, p. 51-65.

Research output: Contribution to journalArticle

@article{ca1c878cd7874832be5457fcb6b58514,
title = "Estimates for cell counts and common odds ratio in three-way contingency tables by homogeneous log-linear models with missing data",
abstract = "Missing observations often occur in cross-classified data collected during observational, clinical, and public health studies. Inappropriate treatment of missing data can reduce statistical power and give biased results. This work extends the Baker, Rosenberger and Dersimonian modeling approach to compute maximum likelihood estimates for cell counts in three-way tables with missing data, and studies the association between two dichotomous variables while controlling for a third variable in 2 × 2 × K tables. This approach is applied to the Behavioral Risk Factor Surveillance System data. Simulation studies are used to investigate the efficiency of estimation of the common odds ratio.",
keywords = "Common odds ratio, Contingency table, Cross-classified data, Log-linear model, Maximum likelihood method, Missing data, Three-way table",
author = "Rochani, {Haresh D.} and Vogel, {Robert L.} and Samawi, {Hani M.} and Linder, {Daniel F}",
year = "2017",
month = "1",
day = "1",
doi = "10.1007/s10182-016-0275-y",
language = "English (US)",
volume = "101",
pages = "51--65",
journal = "AStA Advances in Statistical Analysis",
issn = "1863-8171",
publisher = "Springer Verlag",
number = "1",

}

TY - JOUR

T1 - Estimates for cell counts and common odds ratio in three-way contingency tables by homogeneous log-linear models with missing data

AU - Rochani, Haresh D.

AU - Vogel, Robert L.

AU - Samawi, Hani M.

AU - Linder, Daniel F

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Missing observations often occur in cross-classified data collected during observational, clinical, and public health studies. Inappropriate treatment of missing data can reduce statistical power and give biased results. This work extends the Baker, Rosenberger and Dersimonian modeling approach to compute maximum likelihood estimates for cell counts in three-way tables with missing data, and studies the association between two dichotomous variables while controlling for a third variable in 2 × 2 × K tables. This approach is applied to the Behavioral Risk Factor Surveillance System data. Simulation studies are used to investigate the efficiency of estimation of the common odds ratio.

AB - Missing observations often occur in cross-classified data collected during observational, clinical, and public health studies. Inappropriate treatment of missing data can reduce statistical power and give biased results. This work extends the Baker, Rosenberger and Dersimonian modeling approach to compute maximum likelihood estimates for cell counts in three-way tables with missing data, and studies the association between two dichotomous variables while controlling for a third variable in 2 × 2 × K tables. This approach is applied to the Behavioral Risk Factor Surveillance System data. Simulation studies are used to investigate the efficiency of estimation of the common odds ratio.

KW - Common odds ratio

KW - Contingency table

KW - Cross-classified data

KW - Log-linear model

KW - Maximum likelihood method

KW - Missing data

KW - Three-way table

UR - http://www.scopus.com/inward/record.url?scp=84978745629&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84978745629&partnerID=8YFLogxK

U2 - 10.1007/s10182-016-0275-y

DO - 10.1007/s10182-016-0275-y

M3 - Article

AN - SCOPUS:84978745629

VL - 101

SP - 51

EP - 65

JO - AStA Advances in Statistical Analysis

JF - AStA Advances in Statistical Analysis

SN - 1863-8171

IS - 1

ER -