Assessment of Composite Indicators Using the ANOVA Model Combined with Multivariate Methods |
| |
Authors: | Cristina Davino Rosaria Romano |
| |
Affiliation: | 1. University of Macerata, piazza Oberdan 3, 62100, Macerata, Italy 2. University of Calabria, Arcavacata di Rende, 87036, Cosenza, Italy
|
| |
Abstract: | The synthesis of a number of single observed indicators into a unique composite indicator involves various subjective choices related, for instance, to the type of combination (linear, non-linear) and to the aggregation method (simple average, geometric average) used in its construction. Thus, it is clearly important to analyse the variability of a composite indicator according to all possible alternatives before employing it in any decision-making process. Within such a framework, in this paper, we present a new approach based on a combination of explorative and confirmative analyses aiming to investigate the impact of different subjective choices on the variability of composite indicators. This new approach also allows the analysis of the related individual differences among the statistical units and the use of external information on the same units to enhance the interpretation of the final results. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|