656 – Section on Statistics and Marketing - 2
A Framework for Modeling Ordinal Data in Rating Surveys
Maria Iannario
University of Naples Federico II
Domenico Piccolo
University of Naples Federico II
In this paper, we generalize a mixture model proposed for ordinal data by specifying subjects' covariates for each component of the mixture. Then, the EM procedure for obtaining the maximum likelihood estimates by means of an effective iterative algorithm has been derived and the asymptotic variance-covariances matrix of parameter estimators has been carried out: these results allow an asymptotically efficient inference. In this respect, we present a step-by-step algorithm in order to perform an effective programming for the implementation of a related software. In addition, we checked the usefulness of the approach by means of a real case study. Some final considerations for future developments conclude the paper.