Performance of logistic regression modeling: beyond the number of events per variable, the role of data structure

Courvoisier, Delphine; Combescure, Christophe; Agoritsas, Thomas; Gayet-Ageron, Angèle; Perneger, Thomas

doi:10.1016/j.jclinepi.2010.11.012

Advanced search

Browse by...

Deposit

Highlights

More informations

Title		Performance of logistic regression modeling: beyond the number of events per variable, the role of data structure
Authors		Courvoisier, Delphine Combescure, Christophe Agoritsas, Thomas Gayet-Ageron, Angèle Perneger, Thomas
Published in		Journal of Clinical Epidemiology. 2011, vol. 64, no. 9, p. 993-1000
Abstract		Logistic regression is commonly used in health research, and it is important to be sure that the parameter estimates can be trusted. A common problem occurs when the outcome has few events; in such a case, parameter estimates may be biased or unreliable. This study examined the relation between correctness of estimation and several data characteristics: number of events per variable (EPV), number of predictors, percentage of predictors that are highly correlated, percentage of predictors that were non-null, size of regression coefficients, and size of correlations.
Keywords		Bias (Epidemiology) — Computer Simulation — Humans — Logistic Models — Research Design — Statistics as Topic/methods
Identifiers		DOI: 10.1016/j.jclinepi.2010.11.012 PMID: 21411281
Full text		Article (Published version) (1.1 MB) - Limited access to UNIGE
Structures		Faculté de médecine / Section de médecine clinique / Département de médecine
Research group		Epidémiologie clinique (115)
Citation (ISO format)		COURVOISIER, Delphine et al. Performance of logistic regression modeling: beyond the number of events per variable, the role of data structure. In: Journal of Clinical Epidemiology, 2011, vol. 64, n° 9, p. 993-1000. doi: 10.1016/j.jclinepi.2010.11.012 https://archive-ouverte.unige.ch/unige:25409