Proficiency tests in forensic science: A step towards identifying expertise

Forensic examiners have been publicly rebuked for implausible explanations and inconsistent practice and yet, are frequently called to testify about forensic evidence in criminal cases. There is a growing sense of urgency to quantify the ability of examiners before they appear in court as expert witnesses. Perception, judgement and the ability to discriminate between complex patterns are common themes of many forensic disciplines. Building upon the fingerprint expertise research of Tangen, et al. (2011), this thesis explains how a similar framework can be developed into a proficiency paradigm and applied to all forensic evidence that requires comparative judgements. Given the increased use of photographs as court evidence and for verifying identity across society, the experiment in this study is a proficiency test of facial recognition. Recognising unfamiliar faces is challenging and people frequently make errors on matching tasks, even in optimal conditions (Megreya & Burton, 2006). Minimal scientific data exists about the range of ability in face recognition and we tested fifty-three novices (psychology students) to establish baseline performance. Twenty-seven surveillance camera operators were nominated as experts for this experiment, based on their daily exposure to faces and photographs. Group comparisons found no significant difference in overall discrimination accuracy. Additional analysis suggests the benefit of expertise may overcome the own-race effect. Within the novice group, accurate identification of own and other-race faces differed between Asian and Caucasian students. Exposure to different races seems to improve discrimination accuracy for other-race faces