Model Based Composite Indicators: New Developments in Partial Least Squares-Path Modeling for the Building of Different Types of Composite Indicators

Composite indicators (CIs), in the social sciences, are used more and more for measuring very complex phenomena as the poverty, the progress and the well-being. Using an approach Model Based in to build CIs, instead of an approach Data Driven, it is possible to consider the role (formative and reflective) of the manifest variables (MVs) and to model the relationships among the CIs. In this article, we begin introducing structural equation modeling (SEM) as a tool for building Model Based CIs. Secondly, among the several methods developed to estimate SEM, we show Partial Least Squares Path Modeling (PLS-PM), due to the key role that estimation of the latent variables (i.e. the CIs) plays in the estimation process. Moreover, we present some recent developments in PLS-PM for the treatment of non metric data, hierarchical data, longitudinal data and multi-block data. Finally, we demonstrate how these recent developments can strongly help in the building of CIs. It is easy to realize, for example, that as a consequence of considering nominal and ordinal data, the knowledge about a phenomenon synthesized by a CI is considerably extended and improved especially for operational use. In order to highlight the potentiality of the proposed approach, the construction of a CI is discussed. In particular, a CI of Social Cohesion, developed by using European Values Study data, will be described in detail.