Expert facial comparison evidence: Science versus pseudo science

Abstract

Confirming the identity of the accused is a critical component of many criminal trials. However, recent evidence suggests this process is highly error prone and leads to unacceptably high rates of wrongful conviction (Innocence Project, 2015). When photographic identification evidence is ambiguous, facial mapping practitioners may be called upon to make comparisons between images of the culprit and the accused. This practice assumes that the techniques employed are reliable and can be used to assist the court in making identity confirmation decisions. However, previous experimental work in this area has established that many of these techniques are unreliable (Kleinberg, Vanezis & Burton, 2007; Strathie, McNeill & White, 2012). We extend these findings by examining another facial mapping technique that uses gridlines, drawn between face-pairs, as a potential face matching aid (Oxlee, 2007). Results show that a simple side-by-side presentation of face-pairs without gridlines produces most accurate responding. Moreover, the application of the grideline technique increases the likelihood that two different face pairs will judged to be the same. These findings suggest that continuing to admit facial mapping evidence in court is likely to increase, rather than decrease, the incidence of wrongful conviction.

Keywords

• facial comparison
• identification of persons
• identification of persons from photographs