Events and Meetings of Italian Statistical Society, Advances in Latent Variables - Methods, Models and Applications

Font Size: 
A georeferenced IRT approach for the localization of public services
Simone Di Zio, Lara Fontanella

Last modified: 2013-06-16

Abstract


Marketing in the public sector is addressed to listen, inform and facilitate the access to services and, especially, to improve the supply of social services. An important stage in the public sector marketing process is the collection of data, inputs and feedbacks from stakeholders, without which it is not possible to orient the choices in order to offer the citizen/user a better quality of services. In this context, it is important to survey, in a differentiated way over the area of interest, the needs of the citizens; to this end we propose a tool, the "spatial questionnaire", that is a detection instrument that allows the direct collection of data referred to single portions of the territory. Item Response Theory models are often used for the detection of latent variables and the needs of a community can be considered as such. We propose here the adaptation of IRT models to the collection of citizens’ requests in order to measure, in a geographically differentiated way, their levels of needs. The data collected with the spatial questionnaire and the resulting parameter estimates of the georeferenced IRT models become a valuable informative base to guide the localization of public services.

 


Full Text: PDF