Optimal Estimator for Logistic Model with Distribution-free Random Intercept

Tanya P. Garcia, Yanyuan Ma

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Logistic models with a random intercept are prevalent in medical and social research where clustered and longitudinal data are often collected. Traditionally, the random intercept in these models is assumed to follow some parametric distribution such as the normal distribution. However, such an assumption inevitably raises concerns about model misspecification and misleading inference conclusions, especially when there is dependence between the random intercept and model covariates. To protect against such issues, we use a semiparametric approach to develop a computationally simple and consistent estimator where the random intercept is distribution-free. The estimator is revealed to be optimal and achieve the efficiency bound without the need to postulate or estimate any latent variable distributions. We further characterize other general mixed models where such an optimal estimator exists.

Original languageEnglish (US)
Pages (from-to)156-171
Number of pages16
JournalScandinavian Journal of Statistics
Volume43
Issue number1
DOIs
StatePublished - Mar 1 2016

Fingerprint

Distribution-free
Logistic Model
Intercept
Estimator
Clustered Data
Model Misspecification
Consistent Estimator
Mixed Model
Latent Variables
Postulate
Longitudinal Data
Gaussian distribution
Covariates
Logistic model
Model
Estimate

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

@article{c5021bc842a34f468f1e81fd7babc5b2,
title = "Optimal Estimator for Logistic Model with Distribution-free Random Intercept",
abstract = "Logistic models with a random intercept are prevalent in medical and social research where clustered and longitudinal data are often collected. Traditionally, the random intercept in these models is assumed to follow some parametric distribution such as the normal distribution. However, such an assumption inevitably raises concerns about model misspecification and misleading inference conclusions, especially when there is dependence between the random intercept and model covariates. To protect against such issues, we use a semiparametric approach to develop a computationally simple and consistent estimator where the random intercept is distribution-free. The estimator is revealed to be optimal and achieve the efficiency bound without the need to postulate or estimate any latent variable distributions. We further characterize other general mixed models where such an optimal estimator exists.",
author = "Garcia, {Tanya P.} and Yanyuan Ma",
year = "2016",
month = "3",
day = "1",
doi = "10.1111/sjos.12170",
language = "English (US)",
volume = "43",
pages = "156--171",
journal = "Scandinavian Journal of Statistics",
issn = "0303-6898",
publisher = "Wiley-Blackwell",
number = "1",

}

Optimal Estimator for Logistic Model with Distribution-free Random Intercept. / Garcia, Tanya P.; Ma, Yanyuan.

In: Scandinavian Journal of Statistics, Vol. 43, No. 1, 01.03.2016, p. 156-171.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Optimal Estimator for Logistic Model with Distribution-free Random Intercept

AU - Garcia, Tanya P.

AU - Ma, Yanyuan

PY - 2016/3/1

Y1 - 2016/3/1

N2 - Logistic models with a random intercept are prevalent in medical and social research where clustered and longitudinal data are often collected. Traditionally, the random intercept in these models is assumed to follow some parametric distribution such as the normal distribution. However, such an assumption inevitably raises concerns about model misspecification and misleading inference conclusions, especially when there is dependence between the random intercept and model covariates. To protect against such issues, we use a semiparametric approach to develop a computationally simple and consistent estimator where the random intercept is distribution-free. The estimator is revealed to be optimal and achieve the efficiency bound without the need to postulate or estimate any latent variable distributions. We further characterize other general mixed models where such an optimal estimator exists.

AB - Logistic models with a random intercept are prevalent in medical and social research where clustered and longitudinal data are often collected. Traditionally, the random intercept in these models is assumed to follow some parametric distribution such as the normal distribution. However, such an assumption inevitably raises concerns about model misspecification and misleading inference conclusions, especially when there is dependence between the random intercept and model covariates. To protect against such issues, we use a semiparametric approach to develop a computationally simple and consistent estimator where the random intercept is distribution-free. The estimator is revealed to be optimal and achieve the efficiency bound without the need to postulate or estimate any latent variable distributions. We further characterize other general mixed models where such an optimal estimator exists.

UR - http://www.scopus.com/inward/record.url?scp=84958155432&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84958155432&partnerID=8YFLogxK

U2 - 10.1111/sjos.12170

DO - 10.1111/sjos.12170

M3 - Article

AN - SCOPUS:84958155432

VL - 43

SP - 156

EP - 171

JO - Scandinavian Journal of Statistics

JF - Scandinavian Journal of Statistics

SN - 0303-6898

IS - 1

ER -