The paper describes how to use hierarchical models to assess the reliability of and agreement between two or more types of measurement device. The idea is illustrated by fitting a linear model with nested random effects to a set of data that was obtained from the calibration of two samples of extremely low frequency magnetic field meters. The paper focuses on the formulation of a suitable model that accounts for the various aspects of the calibration protocol and the subsequent interpretation of the parameter estimates. The approach is very flexible and can easily be tuned to the various needs arising in the measurement agreement framework. It can be seen as an extension of the common practice of using a one-way random-effects model to retrieve a measure of agreement.
A hierarchical modelling approach for measuring reliability of and agreement between two types of magnetic field dosimeter / Brazzale, Alessandra Rosalba; A., Salvan; S., Roletti. - In: JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES C-APPLIED STATISTICS. - ISSN 0035-9254. - STAMPA. - 53:2(2004), pp. 261-278. [10.1046/j.1467-9876.2003.05172.x]
A hierarchical modelling approach for measuring reliability of and agreement between two types of magnetic field dosimeter
BRAZZALE, Alessandra Rosalba;
2004
Abstract
The paper describes how to use hierarchical models to assess the reliability of and agreement between two or more types of measurement device. The idea is illustrated by fitting a linear model with nested random effects to a set of data that was obtained from the calibration of two samples of extremely low frequency magnetic field meters. The paper focuses on the formulation of a suitable model that accounts for the various aspects of the calibration protocol and the subsequent interpretation of the parameter estimates. The approach is very flexible and can easily be tuned to the various needs arising in the measurement agreement framework. It can be seen as an extension of the common practice of using a one-way random-effects model to retrieve a measure of agreement.
Pubblicazioni consigliate
I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris
