Variables are rarely, if ever, measured without error. In this paper we propose to use the Object-Oriented Bayesian Networks architecture to model measurement errors. A mixed measurement error model is introduced to model the respondent error. Then an Object-Oriented Bayesian network, implementing the model above, is developed to represent how the actually observed values are generated from the original ones. Furthermore, potentialities and possible extensions of such an approach are discussed.
Marella, D., Vicard, P. (2010). Measurement error modelling using object-oriented Bayesian networks. In Atti della 45° Riunione Scientifica della Società Italiana di Statistica. PADOVA : CLEUP.
Measurement error modelling using object-oriented Bayesian networks
MARELLA, Daniela;VICARD, Paola
2010-01-01
Abstract
Variables are rarely, if ever, measured without error. In this paper we propose to use the Object-Oriented Bayesian Networks architecture to model measurement errors. A mixed measurement error model is introduced to model the respondent error. Then an Object-Oriented Bayesian network, implementing the model above, is developed to represent how the actually observed values are generated from the original ones. Furthermore, potentialities and possible extensions of such an approach are discussed.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
