Sankhya: The Indian Journal of Statistics

2000, Volume 62, Series A, Pt. 1, pp. 110--127

ON A GLOBAL SENSITIVITY MEASURE FOR BAYESIAN INFERENCE

By

FABRIZIO RUGGERI, CNR-IAMI, Italy

and

SIVA SIVAGANESAN, University of Cincinnati, USA

SUMMARY. We propose a global sensitivity measure that has a common interpretation irrespective of the context of the problem, or the unit of measurements. We argue that it provides additional insight about the nature and the extent of robustness to deviations from a specified prior. We believe that the additional insight provided by this measure can be useful, in particular, when the {\it range} is large and robustness is thought to be lacking. We also study the asymptotic behavior of this global sensitivity measure and find that its asymptotic behavior is similar to that of the (usual) local sensitivity measure.

AMS (1991) subject classification. Primary 62F35; secondary 62C10.

Key words and phrases. Bayesian robustness, global sensitivity, asymptotics.

Full paper (PDF)

This article in Mathematical Reviews