Probabilistic sensitivity measures as information value

Emanuele Borgonovo, Gordon B. Hazen, Victor Richmond R. Jose, Elmar Plischke

Abstract: Highlights•The conditions under which a probabilistic sensitivity measure can be information value are investigated.•Value of information and probabilistic sensitivity measures are connected using the common rationale framework.•Necessary and sufficient conditions for an important property (nullity-implies-independence) of sensitivity measures are given.