Frequently Customer Satisfaction (CS) surveys in services are carried out with respect to body that supply different services articulated in different contexts. These differences imply that CS surveys cannot be performed as a whole at an agency level, but need to be articulated in specifically devoted sub-surveys for each services. Differences of services and, often, also of customers, imply that CS surveys make use of ad hoc questionnaires that are administered to independent samples from

more »

... rent populations. Nevertheless, the goal of these surveys is to obtain information not only to evaluate each service separately, but also to obtain a valuation regarding the agency/firm as a whole. For this reason it is often desirable to obtain a measure of CS to perform separate as well as global CS evaluation. In this paper two different methods are considered to analyze CS data coming from different populations with the aim to compare and to pool CS measured at service level. By using data from an Italian Public Service, this procedure is applied to analyse the users� satisfaction with services supplied.