Do professional facial image comparison training courses work?

Abstract

Facial image comparison practitioners compare images of unfamiliar faces and decide whether or not they show the same person. Given the importance of these decisions for national security and criminal investigations, practitioners attend training courses to improve their face identification ability. However, these courses have not been empirically validated so it is unknown if they improve accuracy. Here, we review the content of eleven professional training courses offered to staff at national security, police, intelligence, passport issuance, immigration and border control agencies around the world. All reviewed courses include basic training in facial anatomy and prescribe facial feature (or 'morphological') comparison. Next, we evaluate the effectiveness of four representative courses by comparing face identification accuracy before and after training in novices (n = 152) and practitioners (n = 236). We find very strong evidence that short (1-hour and half-day) professional training courses do not improve identification accuracy, despite 93% of trainees believing their performance had improved. We find some evidence of improvement in a 3-day training course designed to introduce trainees to the unique feature-by-feature comparison strategy used by facial examiners in forensic settings. However, observed improvements are small, inconsistent across tests, and training did not produce the qualitative changes associated with examiners' expertise. Future research should test the benefits of longer examination-focussed training courses and incorporate longitudinal approaches to track improvements caused by mentoring and deliberate practice. In the absence of evidence that training is effective, we advise agencies to explore alternative evidence-based strategies for improving the accuracy of face identification decisions.

Fig 1. Summary of professional facial image comparison training course content.

We reviewed the content of eleven professional facial image comparison training courses. All courses teach facial anatomy (e.g. top left), and encourage facial feature comparison (e.g. top centre). Most courses teach principles of photography, such as the effect of subject-to-camera distance on face photographs (e.g. top right). Notably, only two courses mention error-rates, and only one course cites empirical research to support the training material. A full version of this review table is available in supplementary materials (S2 Appendix). Illustration of the eye: Reprinted under a CC BY license with permission from Jane Wankmiller. Ear figure: Adapted from the Good, Bad and Ugly image set [18]; Photography figure: Reprinted under a CC BY license, original copyright 2013 by A. M. Burton.

Fig 2. Example stimuli from the face identification tasks.

Example stimuli from the (A) GFMT, (B) GBU tests and feature rating task, (C) the High-to-Low image quality test, and representative stimuli from the (D) casework test and (E) inversion test. In each test, participants were asked to decide if the photos show the same person or different people. The answers to these pairs can be found in the Acknowledgements.

Fig 3. Identification accuracy before and after completing Course A or Course B.

Accuracy on the GFMT, GBU-unlimited and GBU-4 seconds tests before (pre-training) and after (post-training) completing online Courses A or B, or the control training. Across all three tests there was no improvement from training. Error bars show within-subjects corrected standard error of the mean [21].

Fig 4. Identification accuracy of Course C trainees before and after training.

Accuracy on the GFMT, High-to-Low and High-to-High image quality tests for 204 Course C trainees from an Australian government agency and 42 control participants before (pre-training) and after training (post-training). Across all three tests there was no improvement from training. Error bars show within-subjects corrected standard error of the mean [21].

Fig 5. Identification accuracy of Course D trainees before and after training.

Accuracy on the GFMT (A) and casework test (B) at pre- and post-training for 32 Course D trainees from a UK police service and 20 control participants. Course D trainees showed significant improvement on the GFMT but not the casework test. Error bars show within-subjects corrected standard error of the mean [21].

Fig 6. Qualitative measures of facial image comparison expertise before and after training.

(A) The extent to which facial feature similarity ratings predict same/different identities (AUC) for Course D trainees from a UK police service and control participants at pre- and post-training. (B) Accuracy on the inversion test at pre- and post-training for Course D trainees and control participants. Course D trainees and control participants show equivalent feature diagnosticity and inversion effects, indicating that Course D did not produce the qualitative indicators of facial image comparison expertise. Error bars show within-subjects corrected standard error of the mean [21].