www.uhasselt.be
DSpace

Document Server@UHasselt >
Research >
Research publications >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/11253

Title: Estimating precision, repeatability, and reproducibility from Gaussian and non- Gaussian data: a mixed models approach
Authors: Assam Nkouibert, Pryseley
Mintiens, Koen
Knapen, Katia
Van de Stede, Yves
Molenberghs, Geert
Issue Date: 2010
Publisher: ROUTLEDGE JOURNALS, TAYLOR & FRANCIS LTD
Citation: JOURNAL OF APPLIED STATISTICS, 37 (10). p. 1729-1747
Abstract: Quality control relies heavily on the use of formal assessment metrics. In this paper, for the context of veterinary epidemiology, we review the main proposals, precision, repeatability, reproducibility, and intermediate precision, in agreement with ISO (international Organization for Standardization) practice, generalize these by placing them within the linear mixed model framework, which we then extend to the generalized linear mixed model setting, so that both Gaussian as well as non-Gaussian data can be employed. Similarities and differences are discussed between the classical ANOVA (analysis of variance) approach and the proposed mixed model settings, on the one hand, and between the Gaussian and non-Gaussian cases, on the other hand. The new proposals are applied to five studies in three diseases: Aujeszky's disease, enzootic bovine leucosis (EBL) and bovine brucellosis. The mixed-models proposals are also discussed in the light of their computational requirements.
Notes: Molenberghs, G (reprint author)[Pryseley, Assam; Molenberghs, Geert] Univ Hasselt, Interuniv Inst Biostat & Stat Bioinformat, B-3590 Diepenbeek, Belgium. [Pryseley, Assam; Molenberghs, Geert] Katholieke Univ Leuven, Louvain, Belgium. [Mintiens, Koen; Knapen, Katia; Van der Stede, Yves] VAR CODA CERVA, Vet & Agrochem Res Ctr, Brussels, Belgium. geert.molenberghs@uhasselt.be
URI: http://hdl.handle.net/1942/11253
DOI: 10.1080/02664760903150706
ISI #: 000282026800009
ISSN: 0266-4763
Category: A1
Type: Journal Contribution
Validation: ecoom, 2011
Appears in Collections: Research publications

Files in This Item:

Description SizeFormat
Published version216.37 kBAdobe PDF
Peer-reviewed author version288.15 kBAdobe PDF

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.