Green, M. J., Medley, Graham and Browne, W. J.. (2009) Use of posterior predictive assessments to evaluate model fit in multilevel logistic regression. Veterinary Research, Vol.40 (No.4). Article 30. ISSN 0928-4249

Preview

PDF WRAP_Medley_Posterior_predictive.pdf - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader Download (250Kb)

Abstract

Assessing the fit of a model is an important final step in any statistical analysis, but this is not straightforward when complex discrete response models are used. Cross validation and posterior predictions have been suggested as methods to aid model criticism. In this paper a comparison is made between four methods of model predictive assessment in the context of a three level logistic regression model for clinical mastitis in dairy cattle; cross validation, a prediction using the full posterior predictive distribution and two “mixed” predictive methods that incorporate higher level random effects simulated from the underlying model distribution. Cross validation is considered a gold standard method but is computationally intensive and thus a comparison is made between posterior predictive assessments and cross validation. The analyses revealed that mixed prediction methods produced results close to cross validation whilst the full posterior predictive assessment gave predictions that were over-optimistic (closer to the observed disease rates) compared with cross validation. A mixed prediction method that simulated random effects from both higher levels was best at identifying the outlying level two (farm-year) units of interest. It is concluded that this mixed prediction method, simulating random effects from both higher levels, is straightforward and may be of value in model criticism of multilevel logistic regression, a technique commonly used for animal health data with a hierarchical structure.

Item Type:	Journal Article
Subjects:	Q Science > QA Mathematics H Social Sciences > HA Statistics
Divisions:	Faculty of Science > Life Sciences (2010- ) > Biological Sciences ( -2010)
Library of Congress Subject Headings (LCSH):	Statistics -- Methodology, Bayesian statistical decision theory, Mathematical statistics -- Methodology, Mathematical models -- Evaluation, Logistic regression analysis
Journal or Publication Title:	Veterinary Research
Publisher:	EDP Sciences
ISSN:	0928-4249
Date:	July 2009
Volume:	Vol.40
Number:	No.4
Page Range:	Article 30
Identification Number:	10.1051/vetres/2009013
Status:	Peer Reviewed
Access rights to Published version:	Open Access
Funder:	Wellcome Trust (London, England)
References:	# Browne W.J., Draper D., A comparison of Bayesian and likelihood-based methods for fitting multilevel models, Bayesian Analysis (2006) 1:473–514. # Dohoo I.R., Martin W., Stryhn H., Veterinary epidemiologic research, Atlantic Veterinary College Inc., Prince Edward Island, Canada, 2003. # Gelman A., Meng X., Stern H., Posterior predictive assessment of model fitness via realized discrepancies, Statistica Sinica (1996) 6:733–807. # Gelman A., Goegebeur Y., Tuerlinckx F., van Mechelen I., Diagnostic checks for discreet data regression models using posterior predictive simulations, Appl. Stat. (2000) 49:247–268. # Goldstein H., Multilevel Statistical Models, London, Edward Arnold, 1995. # Green M.J., Bradley A.J., Medley G.F., Browne W.J., Cow, farm and management factors during the dry period that determine the rate of clinical mastitis after calving, J. Dairy Sci. (2007) 90:3764–3776. # Landwehr J.M., Pregibon D., Shoemaker A.C., Graphical methods for assessing logistic regression models (with discussion), J. Am. Stat. Assoc. (1984) 79:61–83. # Langford I., Lewis T., Outliers in multilevel data, J. R. Stat. Soc. Ser. A (1998) 161:121–160. # Lewin A., Richardson S., Marshall C., Glazier A., Aitman T., Bayesian modelling of differential gene expression, Biometrics (2006) 62:1–9. # Marshall E.C., Spiegelhalter D.J., Approximate cross-validatory predictive checks in disease mapping, Stat. Med. (2003) 22:1649–1660. # Rasbash J., Browne W.J., Healy M., Cameron B., Charlton C., MLwiN Version 2.02, Multilevel Models Project, Centre for Multilevel Modelling, Bristol, UK, 2005. # Spiegelhalter D.J., Thomas A., Best N., Win-BUGS Version 1.4.1., Imperial College and MRC, UK, 2004. # Stern H.H., Cressie N., Posterior predictive model checks for disease mapping models, Stat. Med. (2000) 19:2377–2397.
URI:	http://wrap.warwick.ac.uk/id/eprint/1287

Request changes to a record

Actions (login required)

View Item