A comparative study of the bias corrected estimates in logistic regression.
T. Maiti, and V. Pradhan. Statistical methods in medical research, 17 (6):
621-34(December 2008)5470<m:linebreak></m:linebreak>JID: 9212457; 2008/03/28 aheadofprint; 2008/04/03 aheadofprint; 2008/04/15 aheadofprint; ppublish;<m:linebreak></m:linebreak>Regressió logística.
DOI: 10.1177/0962280207084156
Abstract
Logistic regression is frequently used in many areas of applied statistics. The maximum likelihood estimates (MLE) of the logistic regression parameters are usually computed using the iterative Newton-Raphson method. It is well known that these estimates are biased. Several methods are proposed to correct the bias of these estimates. Among them Firth (1993) and Cordeiro and McCullagh (1991) proposed two promising methods. The conditional exact method (CMLE) is popular for small-sample estimates, and is available in many software packages. In this article we compare these methods in terms of their bias. In general, our extensive simulations show that the methods proposed by Cordeiro and McCullagh and by Firth work well, though Cordeiro and McCullagh is slightly better in our simulations. In case of separation, Firth or CMLE can be used; however, a judicious approach is required when there is a wide variation in results. Two real data analyses are given exhibiting these properties. The data analysis also includes bootstrap results.
%0 Journal Article
%1 Maiti2008
%A Maiti, Tapabrata
%A Pradhan, Vivek
%D 2008
%J Statistical methods in medical research
%K Acute Acute:pathology Acute:therapy Bias(Epidemiology) Biometry Biometry:methods CD4-CD8Ratio Databases Factual HIVInfections HIVInfections:blood Humans Infant IntensiveCare IntensiveCare:statistics&numericaldata Leukemia LikelihoodFunctions LogisticModels Myeloid SurvivalAnalysis
%N 6
%P 621-34
%R 10.1177/0962280207084156
%T A comparative study of the bias corrected estimates in logistic regression.
%U http://www.ncbi.nlm.nih.gov/pubmed/18375454
%V 17
%X Logistic regression is frequently used in many areas of applied statistics. The maximum likelihood estimates (MLE) of the logistic regression parameters are usually computed using the iterative Newton-Raphson method. It is well known that these estimates are biased. Several methods are proposed to correct the bias of these estimates. Among them Firth (1993) and Cordeiro and McCullagh (1991) proposed two promising methods. The conditional exact method (CMLE) is popular for small-sample estimates, and is available in many software packages. In this article we compare these methods in terms of their bias. In general, our extensive simulations show that the methods proposed by Cordeiro and McCullagh and by Firth work well, though Cordeiro and McCullagh is slightly better in our simulations. In case of separation, Firth or CMLE can be used; however, a judicious approach is required when there is a wide variation in results. Two real data analyses are given exhibiting these properties. The data analysis also includes bootstrap results.
%@ 0962-2802; 0962-2802
@article{Maiti2008,
abstract = {Logistic regression is frequently used in many areas of applied statistics. The maximum likelihood estimates (MLE) of the logistic regression parameters are usually computed using the iterative Newton-Raphson method. It is well known that these estimates are biased. Several methods are proposed to correct the bias of these estimates. Among them Firth (1993) and Cordeiro and McCullagh (1991) proposed two promising methods. The conditional exact method (CMLE) is popular for small-sample estimates, and is available in many software packages. In this article we compare these methods in terms of their bias. In general, our extensive simulations show that the methods proposed by Cordeiro and McCullagh and by Firth work well, though Cordeiro and McCullagh is slightly better in our simulations. In case of separation, Firth or CMLE can be used; however, a judicious approach is required when there is a wide variation in results. Two real data analyses are given exhibiting these properties. The data analysis also includes bootstrap results.},
added-at = {2023-02-03T11:44:35.000+0100},
author = {Maiti, Tapabrata and Pradhan, Vivek},
biburl = {https://www.bibsonomy.org/bibtex/23c673d7f863022dc0077efd2783b1cd7/jepcastel},
city = {Department of Statistics, Iowa State University, IA, USA.},
doi = {10.1177/0962280207084156},
interhash = {bf05b445ab7fedccdbc1fb448959e10f},
intrahash = {3c673d7f863022dc0077efd2783b1cd7},
isbn = {0962-2802; 0962-2802},
issn = {0962-2802},
journal = {Statistical methods in medical research},
keywords = {Acute Acute:pathology Acute:therapy Bias(Epidemiology) Biometry Biometry:methods CD4-CD8Ratio Databases Factual HIVInfections HIVInfections:blood Humans Infant IntensiveCare IntensiveCare:statistics&numericaldata Leukemia LikelihoodFunctions LogisticModels Myeloid SurvivalAnalysis},
month = {12},
note = {5470<m:linebreak></m:linebreak>JID: 9212457; 2008/03/28 [aheadofprint]; 2008/04/03 [aheadofprint]; 2008/04/15 [aheadofprint]; ppublish;<m:linebreak></m:linebreak>Regressió logística},
number = 6,
pages = {621-34},
pmid = {18375454},
timestamp = {2023-02-03T11:44:35.000+0100},
title = {A comparative study of the bias corrected estimates in logistic regression.},
url = {http://www.ncbi.nlm.nih.gov/pubmed/18375454},
volume = 17,
year = 2008
}