The Nikumaroro Bones: How Can Forensic Scientists Assist Factfinders?

Abstract

Criminalistics compare traces of unknown origin from crime scenes to traces from specimens of known origin to help determine whether the two specimens have a common origin. They typically testify by giving a firm opinion as to whether a suspect is or is not the source of the trace. This Article describes an alternative to source attribution that has a strong following among students of inference in forensic science. It can be called an evidentiary-value approach that uses relative “likelihoods” or “support” to indicate the weight or probative value of evidence. Essentially, it consists of describing the noteworthy features being compared and explaining how strongly the observations of these features support competing conclusions. To underscore and clarify the differences between the traditional, conclusion-centric approach and the newer evidence-centric perspective, the article uses a recent analysis of data from bones that might be those of the celebrated aviator, Amelia Earhardt, to distinguish between data from the specimens, inferences from the data (conclusions), and statements about the probativity of the data with respect to those inferences.

Keywords: trace evidence, probative value, likelihood, Bayes, statistical inference, forensic anthropology, forensic science

JEL Classification: C11