|
Abstract: Part I of this article described some classic cases in which dubious testimony about probabilities helped convict defendants. At the same time, the author suggested that there are circumstances in which “probability evidence” should be admissible, and promised to chart a middle course between Procrustean exclusion of probability evidence and uncritical acceptance of numerobabble. This article discusses a series of related propositions about criminal cases that include “trace evidence” – bloodstains, semen, glass fragments, fibers, or other materials – linking a defendant to a crime. The author maintains that reasonable estimates of pertinent population proportions should be admissible, that argument on the part of counsel as to corresponding estimates of the probability of a coincidental misidentification should be permitted, and that neither expert opinions as to whether the defendant left the trace evidence nor displays of the posterior probability that defendant did so should be admissible.
Keywords: Probability Evidence, Scientific Evidence, Criminal Law
|