Warnings and Expert Opinions An Evaluation Methodology based on Fuzzy Probabilities

Abstract

A number of mandated and voluntary standards and guidelines expressed as good practice have been set out for the design of warnings. However, the question always arises as to whether or not a given warning will accomplish (or would have accomplished) its purpose of preventing injury whether or not it follows such guidelines. The answer to this question must be phrased in probabilities and sometimes only in qualitative form such as “low probability”, “high probability”, “more probable than not”, etc. In order to obtain such answers, experts are often consulted for their opinions. A methodology is presented which can be used as a basis for checking the consistency of the final conclusions or opinions using the concept of “fuzzy probabilities” and conceptually simple computations. The methodology is also of use to the expert in formulating his opinion rationally and deducing its implications clearly. This methodology is presented here in the context of the opined probability of effectiveness of warnings and instructions although it may be used in any context in which the total proposition can be phrased as a set of interrelated sub propositions as is common in reliability theory, decision theory, etc.