CoChat
Sign Up
Cited by 112Face Recognition and PerceptionDeception detection and forensic psychologyFace recognition and analysis

The DOI matches its Crossref record; 43 of the 43 references it cites were found in a registry or an index; 41 of the 85 sources its claims cite were checked for support.

Do professional facial image comparison training courses work?

Open access
Towler, Alice, Kemp, Richard I., Burton, A. Mike, Dunn, James D., Wayne, Tanya, Moreton, Reuben, White, David·2019·PLOS ONE
DOI 10.1371/journal.pone.0211037PMC PMC6373902
Open this paper in CoChat — free accountOpen PDF

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.

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

This paper reviews eleven professional facial image comparison training courses and empirically evaluates four representative courses. It finds that short courses do not improve face identification accuracy, while a 3-day course produces small, inconsistent improvements without the qualitative markers of examiner expertise.

Key findings

  • Across 11 reviewed courses, all taught facial anatomy and encouraged facial feature or “morphological” comparison; only two courses mentioned error-rates and only one cited empirical research to support training material.
  • Short courses A, B, and C did not improve accuracy: online Courses A/B showed no significant Training × Test interactions (all F < 1; ηp² = .00–.02), with Bayesian evidence 33–111 times more likely under the null; half-day Course C also showed no significant interactions (GFMT: F(1, 244) = 1.48, p > .05, ηp² = .01; others F < 1), with Bayesian evidence 5–14 times more likely under the null.
  • Despite no accuracy improvement, 93% of Course C trainees believed training had improved their identification accuracy, compared with 31% of controls.
  • The 3-day Course D showed significant improvement on a modified GFMT (F(1, 47) = 4.69, p < .05, ηp² = .09; post-training accuracy 93%) and on the inversion test (F(1, 48) = 4.11, p < .05, ηp² = .08), but not on the casework test (F < 1, ηp² = .02; post-training accuracy 72% vs. 71% for controls) and did not produce qualitative expertise changes (feature-rating AUC interaction F < 1; three-way inversion effect F < 1).

Methods

  • Content review: 11 facial reviewer training courses from government agencies in Australia, Finland, the UK, and USA plus a private biometrics company were classified against FISWG guidelines plus additional criteria, totaling 86 criteria, by two raters as “not at all,” “briefly,” or “in detail.”
  • Training evaluations: pre- to post-test designs with control groups. Online Courses A/B used 90 Australian undergraduate psychology students (30 per condition); half-day Course C used 204 Australian government facial image comparison practitioners (mean 7 years in role) plus 42 Amazon Mechanical Turk controls; 3-day Course D used 32 UK police/civilian trainees plus 20 British undergraduate controls.
  • Tests included the Glasgow Face Matching Test, GBU-unlimited and GBU-4 seconds, High-to-Low and High-to-High image quality tests, a casework test, a feature rating task, and an inversion test. Accuracy was analyzed with ANOVA using Training as a between-subjects factor and Test as a within-subjects factor, focusing on the Training × Test interaction, with complementary Bayesian analyses.

Limitations

  • The paper states that the content review did not have access to course material ordinarily delivered verbally by instructors, and that the evaluation approach was not designed to capture mentoring, feedback, deliberate practice, or longitudinal on-the-job learning.
  • The paper states that some Course D data files were corrupt and could not be analyzed, with exact sample sizes per test in supplementary materials; it also notes that Course D improvements were small, inconsistent across tests, and not accompanied by qualitative markers of facial examiner expertise.
  • Evident from the text: only four of the eleven reviewed courses were empirically evaluated; control groups differed from trainees in age, occupation, and recruitment source (undergraduates or MTurk workers vs. practitioners), and there was no long-term follow-up to track training or on-the-job development.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0211037

Cites

43 of 43 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

41 of 85 cited sources checked · 26 supported · 2 flagged · accuracy not yet measured

  • Premisecitation

    Recruiting people with natural aptitude in face identification tasks has been shown to produce substantial boosts to accuracy.

    “Recruiting people with natural aptitude in face identification tasks [37, 38]”
    Quote in the paperNot a substantial claim2 of 2 sources checked
    • cites [37] T Balsdon, 2018, Improving face identification with specialist teams, Cognitive Research: Principles and Implications source

      The Discussion reports that selecting high performers using face-identification tests improved accuracy by about 7%, a gain the authors describe as modest rather than substantial. The Background also cautions that the stronger performance of existing specialist teams cannot be attributed to recruitment alone because their selection and training processes are not documented. (LLM, conf=0.88)

      Checked by a reasoning model — may be wrong.

      Experimental — accuracy not yet measured
    • cites [38] AK Bobak, 2016, Solving the border control problem: Evidence of enhanced face matching in individuals with extraordinary face recognition skills, PLoS ONE source
  • Premisecitation

    None of these studies included facial examiners.

    “However, none of these studies included facial examiners.”
    Quote in the paperSubstantial claim3 of 4 sources checked
    • cites [9] D White, 2014, Passport officers’ errors in face matching, PLoS ONE source
    • cites [10] D White, 2015, Error rates in users of automatic face recognition software, PLoS One source

      Experiment 2 (Method) explicitly includes seven specialist Facial Examiners as participants. Its Results compare their performance with facial review staff and students. Thus, if this paper is one of “these studies,” the claim that none included facial examiners is contrary to our copy. (LLM, conf=0.89)

      Checked by a reasoning model — may be wrong.

      Experimental — accuracy not yet measured
    • cites [27] MH Papesh, 2018, Photo ID verification remains challenging despite years of practice, Cognitive Research: Principles and Implications source

      The Conclusions (page unavailable) identify Papesh’s participants as notaries, bank tellers, and undergraduate students—not facial examiners. That supports the statement for this study, but the paper cannot establish that none of the other studies meant by “these studies” included examiners; its Background describes a separate study that did include forensic facial identification examiners. (LLM, conf=0.78)

      Experimental — accuracy not yet measured
    • cites [28] BE Wirth, 2017, An easy game for frauds? Effects of professional experience and time pressure on passport-matching performance, Journal of Experimental Psychology: Applied source
  • Premisecitation

    In previous work, examiners' accuracy was less impaired when face images were presented upside down compared to novices.

    “and their accuracy was less impaired when face images were presented upside down compared to novices [20, 22].”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [20] A Towler, 2017, Evaluating the feature comparison strategy for forensic face identification, Journal of Experimental Psychology: Applied source
    • cites [22] D White, 2015, Perceptual expertise in forensic facial image comparison, Proceedings of the Royal Society of London B: Biological Sciences source
  • Interpretationcitation

    These possibilities do not account for the qualitative differences observed in their performance in previous work.

    “However, these possibilities do not account for the qualitative differences observed in their performance in previous work [20, 22].”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [20] A Towler, 2017, Evaluating the feature comparison strategy for forensic face identification, Journal of Experimental Psychology: Applied source
    • cites [22] D White, 2015, Perceptual expertise in forensic facial image comparison, Proceedings of the Royal Society of London B: Biological Sciences source

      The Discussion (page unavailable) identifies qualitative differences in forensic examiners’ performance and says they suggest a contribution from differences in cognitive processing. It also says motivation alone has limited explanatory power, but does not establish that all the unspecified ‘possibilities’ cannot account for those differences. (LLM, conf=0.78)

      Experimental — accuracy not yet measured
  • Premisecitation

    The superior accuracy experienced by facial examiners is typically accompanied by qualitative differences in their performance.

    “However, the superior accuracy experienced by facial examiners is typically accompanied by qualitative differences in their performance.”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [20] A Towler, 2017, Evaluating the feature comparison strategy for forensic face identification, Journal of Experimental Psychology: Applied source
    • cites [22] D White, 2015, Perceptual expertise in forensic facial image comparison, Proceedings of the Royal Society of London B: Biological Sciences source
  • It may state two separate claims as one. 28
  • Probably not a substantial claim. 28

  • It may state two separate claims as one.

    Premisecitation

    In a study where passport officers searched for a target person in an array of faces selected by facial recognition software, passport officers made errors on more than 50% of decisions-the same rate as untrained novices.

    “Passport officers made errors on more than 50% of decisions—the same rate as untrained novices.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [10] D White, 2015, Error rates in users of automatic face recognition software, PLoS One source

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

    Premisecitation

    Facial examiners are a specialist group of facial image comparison practitioners who conduct slow, systematic, feature-based comparisons, often as part of criminal investigations.

    “Facial examiners are a specialist group of facial image comparison practitioners who conduct slow, systematic, feature-based comparisons, often as part of criminal investigations (see [20, 22, 23]).”
    Quote in the paperNot a substantial claim2 of 3 sources checked
    • cites [20] A Towler, 2017, Evaluating the feature comparison strategy for forensic face identification, Journal of Experimental Psychology: Applied source
    • cites [22] D White, 2015, Perceptual expertise in forensic facial image comparison, Proceedings of the Royal Society of London B: Biological Sciences source
    • cites [23] PJ Phillips, 2018, Face recognition accuracy in forensic examiners, super-recognisers and algorithms, Proceedings of the National Academy of Sciences source

      The paper describes forensic facial examiners as extensively trained professionals whose image comparisons are rigorous and time-consuming, and whose decisions can support prosecutions and court testimony. Materials and Methods says they used their laboratory tools and methods. Our copy does not establish that their comparisons are specifically feature-based, so that part of the attributed description is unsupported. (LLM, conf=0.84)

      Experimental — accuracy not yet measured

  • It may state two separate claims as one.

    Premisecitation

    In a live photo-to-person identity verification task, passport officers accepted 14% of fraudulent ID documents and rejected 6% of valid IDs.

    “In a live photo-to-person identity verification task for example, passport officers accepted 14% of fraudulent ID documents and rejected 6% of valid IDs [9].”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [9] D White, 2014, Passport officers’ errors in face matching, PLoS ONE source

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Interpretationreasoning

    Differences in the availability of ongoing on-the-job training and opportunities for deliberate practice may explain why facial examiners tend to outperform facial reviewers.

    “Differences in the availability of ongoing on-the-job training and opportunities for deliberate practice (see [29]) may explain why facial examiners tend to outperform facial reviewers [10, 23].”
    Quote in the paperNot a substantial claim2 of 3 sources checked
    • cites [29] KA Ericsson, 1993, The role of deliberate practice in the acquisition of expert performance, Psychological Review source
    • cites [10] D White, 2015, Error rates in users of automatic face recognition software, PLoS One source
    • cites [23] PJ Phillips, 2018, Face recognition accuracy in forensic examiners, super-recognisers and algorithms, Proceedings of the National Academy of Sciences source
    • Rests on an assumption: Facial reviewers do not typically have the same opportunities for mentoring, feedback and extensive practice as facial examiners, and tend to rely completely on training courses like those reviewed here.

  • It may state two separate claims as one.

    Premisecitation

    Mentors provide examiners with feedback on the accuracy of their decisions and feature-based analysis, which may also facilitate learning.

    “Similarly, mentors provide examiners with feedback on the accuracy of their decisions and feature-based analysis, which may also facilitate learning [19].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [19] D White, 2014, Feedback training for facial image comparison, Psychonomic Bulletin and Review source

      The Experiment 1 section describes trial-by-trial feedback on face-matching decisions, and the General Discussion reports that feedback training improved accuracy and may have directed attention to diagnostic facial features. The paper does not describe mentors providing feedback to examiners on their feature-based analyses; that setting and form of feedback are not established by the study. (LLM, conf=0.88)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim.

    Premisecitation

    Previous studies have examined the relationship between the length of time employed in a professional role and face identification accuracy to understand the role of on-the-job learning.

    “However, previous studies have examined the relationship between the length of time employed in a professional role and face identification accuracy to understand the role of on-the-job learning [9, 10, 27, 28].”
    Quote in the paperSubstantial claim3 of 4 sources checked
    • cites [9] D White, 2014, Passport officers’ errors in face matching, PLoS ONE source
    • cites [10] D White, 2015, Error rates in users of automatic face recognition software, PLoS One source
    • cites [27] MH Papesh, 2018, Photo ID verification remains challenging despite years of practice, Cognitive Research: Principles and Implications source
    • cites [28] BE Wirth, 2017, An easy game for frauds? Effects of professional experience and time pressure on passport-matching performance, Journal of Experimental Psychology: Applied source

  • It may state two separate claims as one.

    Premisecitation

    These studies all show no improvement with more professional experience, suggesting that on-the-job learning in these particular roles does not help develop expertise.

    “These studies all show no improvement with more professional experience, suggesting that on-the-job learning in these particular roles does not help develop expertise.”
    Quote in the paperNot a substantial claim3 of 4 sources checked
    • cites [9] D White, 2014, Passport officers’ errors in face matching, PLoS ONE source
    • cites [10] D White, 2015, Error rates in users of automatic face recognition software, PLoS One source

      Experiment 2 Results report no relationship between years employed and face-matching accuracy for either facial review staff or facial examiners. But the General Discussion says examiners may have developed better strategies through feedback and working with specialists, so the finding does not establish that on-the-job learning does not help develop expertise. (LLM, conf=0.85)

      Experimental — accuracy not yet measured
    • cites [27] MH Papesh, 2018, Photo ID verification remains challenging despite years of practice, Cognitive Research: Principles and Implications source
    • cites [28] BE Wirth, 2017, An easy game for frauds? Effects of professional experience and time pressure on passport-matching performance, Journal of Experimental Psychology: Applied source

  • It may state two separate claims as one.

    Findingown data

    Course D trainees achieved just 72% accuracy on the casework test after training, comparable to novice control participants who achieved 71%.

    “Performance on the casework test was also notably low, with Course D trainees achieving just 72% accuracy after training—comparable to novice control participants who achieved 71%.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Findingown data

    Genuine trainees (n = 32) showed small but significant improvements in accuracy on a modified version of the GFMT and the inversion test, but not on the casework test.

    “Genuine trainees (n = 32) showed small but significant improvements in accuracy on a modified version of the GFMT and the inversion test, but not on the casework test.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The casework test most closely approximated the task trainees are required to perform in daily work.

  • It may state two separate claims as one.

    Findingown data

    Courses A and B do not improve face identification accuracy.

    “providing no evidence that Courses A and B improve face identification accuracy.”
    • Also stated in: § Online training course evaluation · boosted
      “These analyses provide between ‘very strong’ and ‘extreme’ evidence in support of the null hypothesis, i.e. that Courses A and B do not improve face identification accuracy.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Findingcitation

    All reviewed courses include practical exercises and provide feedback on the accuracy of facial comparison decisions to trainees.

    “All courses include practical exercises and provide feedback on the accuracy of facial comparison decisions to trainees (see [19]).”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Findingown data

    The improvements observed in Course D were small, inconsistent across tests and not accompanied by qualitative markers of expertise in facial examination.

    “The improvements observed in Course D were small, inconsistent across tests and not accompanied by qualitative markers of expertise in facial examination.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Findingcitation

    Course D did not produce the qualitative changes associated with expertise in facial image comparison: increased diagnosticity of facial feature similarity ratings and a reduced inversion effect.

    “In addition, Course D did not produce the qualitative changes associated with expertise in facial image comparison: increased diagnosticity of facial feature similarity ratings and a reduced inversion effect [20, 22].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [20] A Towler, 2017, Evaluating the feature comparison strategy for forensic face identification, Journal of Experimental Psychology: Applied source
    • cites [22] D White, 2015, Perceptual expertise in forensic facial image comparison, Proceedings of the Royal Society of London B: Biological Sciences source

      Checked before — its earlier verdict is held back until a new check runs.

  • It may state two separate claims as one.

    Findingown data

    Our results show that 93% of Course C trainees believed they had improved as a result of training, despite this training not improving accuracy.

    “This is especially important because our results show that 93% of Course C trainees believed they had improved as a result of training, despite this training not improving accuracy.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Study design and conductnone given

    The casework test is representative of forensic facial examination casework.

    “The casework test is representative of forensic facial examination casework and was supplied by Course D’s host agency (see Methods for details).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may state two separate claims as one.

    Interpretationreasoning

    FISWG guidelines do not make reference to empirical research supporting the validity of the recommended training practices, so adherence to the guidelines provides no evidence that a particular training course is effective.

    “The FISWG guidelines do not make reference to empirical research supporting the validity of the recommended training practices, so adherence to the guidelines provides no evidence that a particular training course is effective.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim.

    Interpretationnone given

    This paper provides the first comprehensive evaluation of the effectiveness of professional facial image comparison training courses.

    “This paper provides the first comprehensive evaluation of the effectiveness of professional facial image comparison training courses.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

    Premisecitation

    Short training courses are used routinely by government agencies around the world and comply with FISWG training guidelines.

    “This is despite short courses being used routinely by government agencies around the world and complying with FISWG training guidelines (see S1 Appendix).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites S1 Appendix

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • It may state two separate claims as one.

    Interpretationreasoning

    The courses evaluated here were representative of accepted practice in facial image comparison training, and so results are likely to be indicative of the effectiveness of professional training courses more generally.

    “The courses evaluated here were representative of accepted practice in facial image comparison training, and so results are likely to be indicative of the effectiveness of professional training courses more generally.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Interpretationcitation

    These qualitative differences appear to reflect a greater reliance on feature-level details when comparing face images.

    “These qualitative differences appear to reflect a greater reliance on feature-level details when comparing face images.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationown data

    People do not have insight into the effectiveness of facial image comparison training, and perceived benefit of training is unrelated to its actual benefit.

    “This suggests that people do not have insight into the effectiveness of facial image comparison training, and that perceived benefit of training is unrelated to its actual benefit.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Study design and conductnone given

    Course D trainees had not previously seen the images in the casework test.

    “Course D trainees had not previously seen the images in the casework test.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Premisecitation

    Professional facial image comparison training courses have not been empirically validated, so it is unknown whether they improve identification accuracy.

    “However, these courses have not been empirically validated so it is unknown if they improve accuracy.”
    • Also stated in: § Background · hedged
      “The only published evaluation of a professional facial image comparison training course was conducted in the 1970’s by psychologists Woodhead, Baddeley, and Simmonds [4].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [4] MM Woodhead, 1979, On training people to recognize faces, Ergonomics source

  • It may state two separate claims as one.

    Premisecitation

    Training is accepted as evidence of expertise in court, allowing 'trained experts' to present highly persuasive identification evidence at trial.

    “More generally, training is also accepted as evidence of expertise in court [11], allowing ‘trained experts’ to present highly persuasive identification evidence at trial [12].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [11] Daubert v. Merrell Dow Pharmaceuticals, Inc.: Supreme Court of the United States; 1993.

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [12] A Towler, 2018, Are forensic scientists experts?, Journal of Applied Research in Memory and Cognition source

      Checked before — its earlier verdict is held back until a new check runs.

  • Probably not a substantial claim.

    Study design and conductnone given

    For all tests, items were split into two sets of equal difficulty using accuracy data from previous studies so that equivalent versions could be administered at pre- and post-test.

    “For all tests, we split items into two sets of equal difficulty using accuracy data from previous studies so that equivalent versions could be administered at pre- and post-test.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. It may state two separate claims as one.

    Study design and conductnone given

    We created two equally difficult 20-item versions (10 match, 10 non-match) of the High-to-High and High-to-Low image quality tests, and allocation of each version to pre- and post-training was counterbalanced across participants.

    “We created two equally difficult 20-item versions (10 match, 10 non-match) of the High-to-High and High-to-Low image quality tests. Allocation of each version to pre- and post-training was counterbalanced across participants.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Study design and conductnone given

    Some trainees' data files were corrupt and could not be analysed.

    “Some trainees’ data files were corrupt and could not be analysed.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Study design and conductnone given

    Test order was randomised for each participant.

    “Test order was randomised for each participant.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim.

    Study design and conductnone given

    Two versions were randomly assigned to the upright and inverted blocks at pre- and post-training for each participant.

    “Two versions were randomly assigned to the upright and inverted blocks at pre- and post-training for each participant.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

    Study design and conductnone given

    The order of the GBU-unlimited and GBU-4 seconds was counterbalanced across participants, but held constant pre- to post-training.

    “The order of the GBU-unlimited and GBU-4 seconds was counterbalanced across participants, but held constant pre- to post-training.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may state two separate claims as one.

    Study design and conductcitation

    Following Towler, White and Kemp [17], the GFMT was split into two equally difficult 20-item versions (10 match, 10 non-match), and allocation of each version to the pre- and post-training was counterbalanced across participants.

    “Following Towler, White and Kemp [17], the GFMT was split into two equally difficult 20-item versions (10 match, 10 non-match). Allocation of each version to the pre- and post-training was counterbalanced across participants.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [17] A Towler, 2014, Evaluating training methods for facial image comparison: The face shape strategy does not work, Perception source

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Study design and conductreasoning

    In an effort to equate motivation between the groups, the highest-performing control participant across the pre- and post-training tests received an additional £50 Amazon voucher.

    “In an effort to equate motivation between the groups, the highest-performing control participant across the pre- and post-training tests received an additional £50 Amazon voucher.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

    Study design and conductnone given

    Task order was randomised for each participant, but held constant pre- to post-training.

    “Task order was randomised for each participant, but held constant pre- to post-training.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may state two separate claims as one.

    Study design and conductcitation

    We created four equally difficult 30-item versions (15 match, 15 non-match) using human performance data from O'Toole, An, Dunlop, and Natu [41], and allocation of each version to the pre- or post-training and GBU-unlimited and GBU-4 seconds was counterbalanced across participants.

    “We created four equally difficult 30-item versions (15 match, 15 non-match) using human performance data from O’Toole, An, Dunlop, and Natu [41]. Allocation of each version to the pre- or post-training and GBU-unlimited and GBU-4 seconds was counterbalanced across participants.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [41] AJ O’Toole, 2012, Comparing face recognition algorithms to humans on challenging tasks, ACM Transactions on Applied Perception source

  • It may state two separate claims as one.

    Premisecitation

    Facial examiners' training also includes extensive on-the-job mentoring from more senior analysts, which may be critical in developing their expertise.

    “Second, facial examiners’ training also includes extensive on-the-job mentoring from more senior analysts, which may be critical in developing their expertise.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [25] AJ Dowsett, 2014, Unfamiliar face matching: Pairs out-perform individuals and provide a route to training, British Journal of Psychology source

      Checked before — its earlier verdict is held back until a new check runs.

  • Probably not a substantial claim. It may state two separate claims as one.

    Study design and conductnone given

    The casework test had not been previously seen by trainees, and we created two equally difficult versions of the test, consisting of 24 trials each (12 match, 12 non-match).

    “The test had not been previously seen by trainees. We created two equally difficult versions of the test, consisting of 24 trials each (12 match, 12 non-match).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

    Study design and conductreasoning

    One version of the casework test was always administered at pre-training, and always as the last test, as it formed the basis of a pre-existing feedback training module.

    “One version of the casework test was always administered at pre-training, and always as the last test, as it formed the basis of a pre-existing feedback training module.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may state two separate claims as one.

    Study design and conductcitation

    The two equally difficult halves of the GFMT used in the previous evaluations were adapted to exclude identities used in the Cambridge Face Memory Test-Australian (see [42]), and test version order was randomly allocated to pre- and post-training for each participant.

    “The two equally difficult halves of the GFMT used in the previous evaluations were adapted to exclude identities used in the Cambridge Face Memory Test—Australian (see [42]). One version of this modified GFMT contained 17-items (9 match, 8 non-match) and the other contained 18-items (10 match, 8 non-match). Test version order was randomly allocated to pre- and post-training for each participant.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [42] E McKone, 2011, Face ethnicity and measurement reliability affect face recognition performance in developmental prosopagnosia: Evidence from the Cambridge Face Memory Test–Australian, Cognitive Neuropsychology source

  • Probably not a substantial claim.

    Study design and conductreasoning

    To equate the groups' experience as much as possible, control participants completed the pre-training tests on Day 1 and mock training on Days 2 and 3.

    “To equate the groups’ experience as much as possible, control participants completed the pre-training tests on Day 1 and mock training on Days 2 and 3.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Study design and conductcitation

    The Glasgow Face Matching Test (GFMT) is a standardised psychometric test of face matching ability.

    “The Glasgow Face Matching Test (GFMT) is a standardised psychometric test of face matching ability [6].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [6] AM Burton, 2010, The Glasgow Face Matching Test, Behavior Research Methods source

  • Probably not a substantial claim. It may state two separate claims as one.

    Study design and conductnone given

    We removed 10 image pairs from the model image set that were used in a pre-existing facial feature analysis exercise during training, and we then created four equally difficult versions, each containing 12 trials (6 match, 6 non-match).

    “We removed 10 image pairs from the model image set that were used in a pre-existing facial feature analysis exercise during training (see S3 Appendix for details of this exercise). We then created four equally difficult versions, each containing 12 trials (6 match, 6 non-match).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Study design and conductown data

    Training courses were classified as having met each criterion either 'not at all', 'briefly' or 'in detail' by two raters, and disagreement between the raters was resolved by joint review of the material.

    “Training courses were classified as having met each criterion either ‘not at all’, ‘briefly’ or ‘in detail’ by two raters. Disagreement between the raters was resolved by joint review of the material.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Study design and conductnone given

    Our approach to evaluating the effectiveness of training courses was not designed to capture these sources of learning.

    “Our approach to evaluating the effectiveness of training courses was not designed to capture these sources of learning.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Study design and conductown data

    Ninety Australian undergraduate psychology students were randomly allocated to one of three conditions, such that 30 participants were in each.

    “Ninety Australian undergraduate psychology students (Mean age = 19 years, SD = 3 years) participated in return for course credit. They were randomly allocated to one of three conditions, such that 30 participants were in each.”
    Quote in the paperNot a substantial claim
44 of 44

  • The paper relies on this without support.

    In many organisations, attending a facial image comparison training course is taken as evidence of competency.

    “In many organisations, attending a facial image comparison training course is taken as evidence of competency.”

  • The paper relies on this without support.

    The courses evaluated here qualify practitioners within their own organisations to make critical identification decisions in police, intelligence and national security operations, at borders, and when issuing secure identity documents.

    “The courses we evaluate here qualify practitioners within their own organisations to make critical identification decisions in police, intelligence and national security operations, at borders, and when issuing secure identity documents (e.g., passports).”

  • The paper relies on this without support.

    It is unclear to what extent agencies adhere to official guidelines for facial image comparison training.

    “but it is unclear to what extent agencies adhere to them.”

  • The paper relies on this without support.

    Very little is known about professional facial image comparison training.

    “However, very little is known about this training.”

  • The paper relies on this without support.

    The casework test most closely approximated the task trainees are required to perform in daily work.

    “the task that most closely approximated the task trainees are required to perform in daily work”

  • The paper relies on this without support.

    Facial reviewers do not typically have the same opportunities for mentoring, feedback and extensive practice as facial examiners, and tend to rely completely on training courses like those reviewed here.

    “Facial reviewers do not typically have the same opportunities for mentoring, feedback and extensive practice as facial examiners, and tend to rely completely on training courses like those reviewed here.”

  • The paper relies on this without support.

    The training courses evaluated in the study are among the formal training courses that facial examiners attend.

    “the formal training courses they attend”
7 of 7

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • There are large discrepancies between informal training courses in improvement of facial examiner expertise.

    “A recent study on the performance of informal training courses on facial comparison suggested that there are large discrepancies between courses in improvement of facial examiner expertise [134].”

    From Forensic Facial Comparison: Current Status, Limitations, and Future Directions (2021) · cites this paper as [134]

    The Abstract and General discussion report that short courses did not improve identification accuracy, while the three-day course produced small, inconsistent improvements. They do not establish large discrepancies between courses in improving examiner expertise; the three-day course did not produce the qualitative changes associated with that expertise. (LLM, conf=0.87)

    Experimental — accuracy not yet measured

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • ConsistentF(1, 244) = 1.48reported p > .05 · recomputed p = .225

    “F(1, 244) = 1.48, p > .05”

  • ConsistentF(1, 139) = 2.17reported p > .05 · recomputed p = .143

    “F(1, 139) = 2.17, p > .05”

  • ConsistentF(1, 47) = 4.69reported p < .05 · recomputed p = .035

    “F(1, 47) = 4.69, p < .05”

  • ConsistentF(1, 47) = 5.47reported p < .05 · recomputed p = .024

    “F(1, 47) = 5.47, p < .05”

  • ConsistentF(1, 48) = 4.11reported p < .05 · recomputed p = .048

    “F(1, 48) = 4.11, p < .05”

  • ConsistentF(1, 48) = 21.42reported p < .001 · recomputed p < .001

    “F(1, 48) = 21.42, p < .001”

  • ConsistentF(1, 48) = 1.35reported p > .05 · recomputed p = .251

    “F(1, 48) = 1.35, p > .05”

7 of 7

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Inversion test has conflicting trial counts

    Minor

    The 3-day course evaluation describes the inversion test as having 24 items, but its Methods section lists 36 trials for the same test.

1 of 1

Citations

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Metadata

Authors
Towler, A., Kemp, R. I., Burton, A. M., Dunn, J. D., Wayne, T., Moreton, R., & White, D.
Affiliations
  • UNSW Sydney (AU)
  • University of York (GB)
  • Metropolitan Police Service (GB)
Corresponding author
Towler, Alice
ORCID
Venue
PLOS ONE
Publisher
Public Library of Science (PLoS)
Published
2019-02-13
Volume
14
Issue
2
Pages
e0211037
ISSN
1932-6203 (ISSN-L)
DOI
10.1371/journal.pone.0211037
PMC
PMC6373902
OpenAlex
W2912087339
Semantic Scholar
82c38addb7da183d2a910c5c688df37c50ce217c
Cited by
111
References
42
Citations by year
2026: 6 · 2025: 11 · 2024: 21 · 2023: 12 · 2022: 14 · 2021: 25 · 2020: 14 · 2019: 8
FWCI
8.01
Type
journal article
Language
en
Open access
Licence
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • White Rose Research Online (University of Leeds, The University of Sheffield, University of York) (acceptedVersion, cc-by)
  • Open Research Online - ORO (The Open University) (acceptedVersion)
  • UNSWorks (University of New South Wales, Sydney, Australia) (submittedVersion, cc-by-nc-nd)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Figshare (submittedVersion, cc-by)
Funders
  • Department of Foreign Affairs and Trade, Australian Government
  • Australian Research Council (LP110100448, LP160101523)
  • FP7 Ideas: European Research Council (323262)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Face Recognition and Perception
  • Social Sciences › Psychology › Social Psychology › Deception detection and forensic psychology
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face recognition and analysis
Keywords
face identification, professional training programs, training effectiveness, comparative morphology, National Security Agency, police, border control, identification accuracy, deliberate practice
MeSH terms
Education, Professional, Face, Female, Humans, Male, Pattern Recognition, Visual, Police

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-11-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review