[Forum SIS] Avviso di seminario

[Francesca De Battisti (francesca.debattisti)]

Martedì 28 febbraio 2012 


Federico Andreis (Università degli Studi di Milano) terrà il seminario dal titolo:


 


“Perspectives of multidimensional extensions of IRT models in customer satisfaction evaluation”


 


presso l'Aula Seminari del Dipartimento di Scienze Economiche, Aziendali e Statistiche della Facoltà di Scienze Politiche dell'Università degli Studi di Milano (via Conservatorio, 7). 


 


Il seminario è programmato alle ore 12.15. 


 


ABSTRACT


The assumption of unidimensionality in the usual IRT setting is often not reasonable. Nonetheless, because of the complexity of multidimensional extensions of classic item response theory models, the majority of IRT researchers would still make use of unidimensional techniques, which might lead to suboptimal inference on the (possibly multidimensional in nature) latent trait of interest. The use of IRT modeling, originally intended for psychometric and ability assessment purposes, has been recently introduced in the field of customer satisfaction evaluation; little has been done, however, towards the extension to multiple dimensions. A short review is presented, in which we try to highlight pros and cons of using complex modeling structures such as multidimensional IRT models. Estimation issues arise, as maximum likelihood based methods involve high-dimensional optimization, and implementation of the necessary algorithm becomes not trivial. Monte Carlo Markov Chain methods are proposed as an alternative, the Metropolis-Hastings within Gibbs approach in particular is investigated, which grants a straightforward way to implement such complex estimations; convergence issues are addressed and diagnostic methods are recalled.


  
Tutti gli interessati sono invitati a partecipare.

 

***********************************************************
Francesca De Battisti
Dipartimento di Scienze Economiche, Aziendali e Statistiche 
Sezione Statistica e Matematica
(III piano, studio n. 29)
Via Conservatorio 7
20122 - Milano
Tel: 02.50321464
Fax: 02.50321505
***********************************************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.stat.unipg.it/pipermail/sis/attachments/20120217/f6f0b3e9/attachment-0001.html>