Article ID: | iaor2009717 |
Country: | United Kingdom |
Volume: | 59 |
Issue: | 2 |
Start Page Number: | 239 |
End Page Number: | 251 |
Publication Date: | Feb 2008 |
Journal: | Journal of the Operational Research Society |
Authors: | Saltelli A., Tarantola S., Cherchye L., Moesen W., Rogge N., Puyenbroeck T. Van, Saisana M., Liska R. |
Keywords: | construction & architecture |
Composite indicators (CIs) are often used for benchmarking countries' performance, but they frequently stir controversies about the unavoidable subjectivity in their construction. Data Envelopment Analysis helps to overcome some key limitations, as it does not need any prior information on either the normalization of sub-indicators or on an agreed unique set of weights. Still, subjective decisions remain, and such modelling uncertainty propagates onto countries' CI scores and rankings. Uncertainty and sensitivity analysis are therefore needed to assess the robustness of the final outcome and to analyse how much each source of uncertainty contributes to the output variance.