In this paper an integrated use of NLPCA (Nonlinear PCA) and multilevel
models for the analysis of satisfaction data is proposed. The basic hypothesis is that observed
ordinal variables describe different aspects of a latent unobservable continuous
variable that depends on covariates connected with individual and contextual features.
NLPCA is used to measure the level of a latent variable and multilevel model is adopted
for detecting individual and enviromental determinants of its level. An application to Eurobarometer
survey data concerning the satisfaction of European citizens for some Services
of General Interest, is carried out.
