| One model, several results: the paradox of the Hosmer-Lemeshow goodness-of-fit test for the logistic regression model. | |
| | |
MedLine Citation:
|
PMID: 11055275 Owner: NLM Status: MEDLINE |
Abstract/OtherAbstract:
|
BACKGROUND: The Hosmer-Lemeshow test, used extensively to assess the fit of the logistic regression model, is performed by several statistical packages. Recent studies have shown some problems in the use of this test when ties are present. These problems were attributed merely to the test implementation. METHODS: We analysed the order of the observations as an alternative explanation of the problem of ties. Using a data-set of 1393 intensive care unit (ICU) patients we performed the Hosmer-Lemeshow test with all possible subjects dispositions. RESULTS: We obtained about one million different P values, ranging from 0.01 to 0.95. DISCUSSION: It is already known that when the Hosmer-Lemeshow goodness-of-fit test is performed with a number of covariate patterns lower than the number of subjects, its result may be inaccurate. We showed that the extent of this problem could be relevant under particular conditions. We also suggest a strategy for estimating the extent of the problem and subsequent interpretation. |
| | |
Authors:
|
G Bertolini; R D'Amico; D Nardi; A Tinazzi; G Apolone |
Related Documents
:
|
10399205 - Testing proportionality in the proportional odds model fitted with gee. 15387245 - Bayesian support vector regression using a unified loss function. 16850435 - A weighted logistic regression model for estimation of recurrence of adenomas. 8044725 - Regression in the service of the superego: the do's and don'ts of stepwise multiple reg... 9178985 - Malar recontouring using 3d bio-modelling. 20085815 - Improved kymography tools and its applications to mitosis. |
Publication Detail:
|
Type: Journal Article |
Journal Detail:
|
Title: Journal of epidemiology and biostatistics Volume: 5 ISSN: 1359-5229 ISO Abbreviation: J Epidemiol Biostat Publication Date: 2000 |
Date Detail:
|
Created Date: 2001-03-09 Completed Date: 2001-05-31 Revised Date: 2007-11-15 |
Medline Journal Info:
|
Nlm Unique ID: 9611181 Medline TA: J Epidemiol Biostat Country: ENGLAND |
Other Details:
|
Languages: eng Pagination: 251-3 Citation Subset: IM |
Affiliation:
|
Clinical Epidemiology Unit, Institute of Research Pharmacology Mario Negri, Ranica, Bergamo, Italy. |
Export Citation:
|
APA/MLA Format Download EndNote Download BibTex |
| MeSH Terms | |
Descriptor/Qualifier:
|
Hospital Mortality Humans Intensive Care Units Italy Logistic Models* Predictive Value of Tests Severity of Illness Index Statistics as Topic / methods |
From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine
Previous Document: Historical HIV prevalence in Edinburgh Prison: a database-linkage study.
Next Document: A comparison of primary and proxy respondent reports of habitual physical activity, using kappa stat...