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

Font Size: 
Composite Indices Building: A Comparison of Unbalance Adjustment Methods
Matteo Mazziotta, Adriano Pareto

Last modified: 2013-06-14

Abstract


In this paper, we analyze some unbalance adjustment methods for building composite indices with a non-compensatory approach. In particular, we empirically compare the Mazziotta-Pareto Index (MPI) and the Mean-Min Function(MMF). The MPI is a not full compensatory index based on the use of a penalty related to the unbalance among dimensions. The MMF allows compensability among dimensions with a cost that increases with unbalance and can be seen as an intermediate case between perfect  substitutability and perfect complementarity among dimensions.

Full Text: PDF