Comments on the relationship between the impact and the reliability of evidence

Abstract In some inference situations a report of an event must be distinguished from the event itself. Such a distinction is necessary whenever there is reason to question the reliability of the source which generates the report. A quantity called adjusted likelihood ratio (A) is defined which incorporates information about source reliability and information about the inferential impact (L) of the event being reported. A is shown to differ from L under several different conditions of source reliability. When source reliability is not contingent upon the truth of any hypothesis being considered, a small reduction in reliability drastically degrades very large values of L. When L is small to begin with the effects of reducing reliability are less drastic. When source reliability is contingent upon some hypothesis, then A can be greater than L for the event.