Is it possible to guarantee that the mere exposure of a subject to a belief elicitation task will not affect the very same beliefs that we are trying to elicit? In this paper, we introduce mechanisms that make it simultaneously strictly dominant for the subject (a) not to acquire any information...