Abstract
To evaluate the worth of uncertain information one must obtain three types of evaluative information: 1) statistical measures of the uncertainty of the information and of its likely occurrence; 2) the decision rule (how the information is used) and measures of the effectiveness for the decision action; and 3) the costs of obtaining the information. The steps in information evaluation are illustrated by an example where rainfall information is used to provide flood warning. The results illustrate how both information uncertainty and the decision rule jointly affect the value of the information and that in some circumstances information can have a negative value. In view of the increasing cost of information, it would seem prudent to utilize information evaluation in actions designed to eliminate the collection of information of low value and to efficiently use the information and data that is collected.
