Expert Panel Opinion and Global Sensitivity Analysis for Composite Indicators

摘要

Composite indicators aggregate multi-dimensional processes into simplified concepts often aiming at underpinning the development of data-driven narratives for policy consumption. Due to methodological issues, doubts are often raised about the robustness of the composite indicators and the significance of the associated policy messages. In this paper we use expert panel information (derived from budget allocation and analytic hierarchy process) on the relative importance of the underlying indicators included in a composite indicator and run in tandem uncertainty and sensitivity analysis to gain useful insights during the process of composite indicators building. We discuss the extent to which variance-based sensitivity analysis may increase transparency or make policy inference more defensible by using the United Nation's Technology Achievement Index as an illustration.