ArticlePDF Available

Self-reported face recognition abilities for own and other-race faces

Authors:

Abstract and Figures

Purpose The other-race effect shows that people are better recognizing faces from their own-race compared to other-race faces. This effect can have dramatic consequences in applied scenarios whereby face identification is paramount, such as eyewitness identification. This paper aims to investigate whether observers have insights into their ability to recognize other-race faces. Design/methodology/approach Chinese ethnic observers performed objective measures of own- and other-race face recognition – the Cambridge Face Memory Test Chinese and the Cambridge Face Memory Test original; the PI20 – a 20-items self-reported measured of general face recognition abilities; and the ORE20 – a new developed 20-items self-reported measure of other-race face recognition. Findings Recognition of own-race faces was better compared to other-race faces. This effect was also evident at a phenomenological level, as observers reported to be worse recognizing other-race faces compared to own-race faces. Additionally, although a moderate correlation was found between own-race face recognition abilities and the PI20, individual differences in the recognition of other-race faces was only poorly associated with observers’ scores in the ORE20. Research limitations/implications These results suggest that observers’ insights to recognize faces are more consistent and reliable for own-race faces. Practical implications Self-reported measures of other-race recognition could produce misleading results. Thus, when evaluating eyewitness’ accuracy identifying other-race faces, objective measures should be used. Originality/value In contrast to own race recognition, people have very limited insights into their recognition abilities for other race faces.
Content may be subject to copyright.
Self-reported face recognition abilities
for own and other-race faces
Alejandro J. Estudillo
Abstract
Purpose The other-race effect shows that people are better recognizing faces from their own-race
compared to other-race faces. This effect can have dramatic consequences in applied scenarios
whereby face identification is paramount, such as eyewitness identification. This paper aims to
investigate whether observers haveinsights into theirability to recognize other-race faces.
Design/methodology/approach Chinese ethnic observers performed objective measures of own-
and other-race face recognition the Cambridge Face Memory Test Chinese and the Cambridge Face
Memory Test original; the PI20 a 20-items self-reported measured of general face recognition abilities;
and the ORE20 a new developed 20-items self-reported measure of other-race face recognition.
Findings Recognition of own-race faces was better compared to other-race faces. This effect was also
evident at a phenomenological level, as observers reported to be worse recognizing other-race faces
compared to own-race faces. Additionally, althougha moderate correlation was found between own-race
face recognition abilities and the PI20, individual differences in the recognition of other-race faces was
only poorly associated with observers’ scores in the ORE20.
Research limitations/implications These results suggest that observers’ insights to recognize faces
are more consistent and reliable for own-race faces.
Practical implications Self-reported measures of other-race recognition could produce misleading
results. Thus, when evaluating eyewitness’ accuracy identifying other-race faces, objective measures
should be used.
Originality/value In contrast to own race recognition, people have very limited insights into their
recognition abilities for other race faces.
Keywords Face recognition, Eyewitness identification, Face identification, Other-race effect, PI20,
Self-reported face recognition
Paper type Research paper
Introduction
People are generally more efficient and accurate recognizing faces of their own-race
compared to other-race faces. This effect, which is known as the other-race effect (ORE,
Malpass and Kravitz, 1969), is evident across different cultures and countries (Meissner
and Brigham, 2001) and has been found with different paradigms, including perceptual
matching (Kokje et al., 2018;Megreya et al.,2011), face recognition (Chiroro and Valentine,
1995;Estudillo et al.,2020;Malpass and Kravitz, 1969;Wong et al., 2020) and line-up
identification tasks (Evans et al., 2009). A meta-analysis study comprising nearly 5000
participants across 39 different studies showed that people are 2.23 times more likely to
recognize own-race faces compared to other-race faces (Meissner and Brigham, 2001).
This evidence points that the ORE is a very robust effect.
Rather than being homogeneous, the size of the ORE presents substantial individual
differences across observers (Wan et al.,2017). For example, a recent study with
Caucasian and Asians observers showed the standard ORE at a group level, that is
recognition performance was better for own compared to other race faces (Wan et al.,
Alejandro J. Estudillo is
based at Department of
Psychology, Bournemouth
University Talbot
Campus, Poole, UK and
Schoool of Psychology,
University of Nottingham,
Selangor, Malaysia.
Received 19 June 2020
Revised 15 August 2020
19 September 2020
Accepted 22 September 2020
Funding: The author received
no specific funding for this
work.
Conflict of interest: The author
declares no conflict of interest.
Ethical Approval: This study
has been run following the
Code of Ethics and Conduct
authorized by American
Psychological Association, and
it was approved by the ethics
committee of the University of
Nottingham Malaysia.
Informed Consent: Participants
gave their informed consent
and were debriefed at the end
of the study.
DOI 10.1108/JCP-06-2020-0025 ©Emerald Publishing Limited, ISSN 2009-3829 jJOURNAL OF CRIMINAL PSYCHOLOGY j
2017). However, an individual differences analysis showed that around 8% of these
observers performed so extremely poor at recognizing other-race faces that they could be
considered to suffer a specific type of face-blindness for other race faces. The authors
concluded that the lack of contact with other-race faces is the main cause of the ORE (Wan
et al.,2017
; see also Estudillo et al., 2020).
Regardless of its origin, it has become clear that the ORE can have catastrophic
consequences in applied scenarios whereby face identification is of paramount importance,
such as in eyewitness identification parades and id-verification settings (e.g. passport
control officers). A paradigmatic case illustrating this issue was Ronald Cotton’s wrongful
conviction case (see www.theinnocenceproject.org). Mr. Cotton is an African-American
citizen who was accused of sexual assault in 1984. The rape victim, a Caucasian lady,
misidentified Mr. Cotton as the rapist. As consequence, Mr. Cotton spent more than
10 years in prison for a crime he did not commit until he was exonerated by DNA evidence
in 1995. Thus, if an eyewitness and the perpetrator are from different races, it is crucial to
determine how reliable an eyewitness is recognizing faces from the perpetrator’s race.
Therefore, developing tools to assess eyewitnesses’ ability to recognize other-race faces is
an important endeavor for forensic scientists.
Recognition confidence (i.e. “how confident are you that you have done a correct recognition
decision?”) is a potential maker that can provide some hints about an eyewitness’s accuracy
identifying a perpetrator. Recognition confidence has been widely used in forensic research
and even police and lawyers consider that this marker can provide a reliable measure of
identification skills (Potter and Brewer, 1999;Sauer et al., 2019). These conclusions are
somehow supported by some research. For example, it has been found that recognition
confidence is positively associated with face identification accuracy (Wixted and Wells,
2017), but this association is modulated by participants’ face recognition abilities. This is such
that poor face recognizers are much more likely to make high confidence identification errors
compared to good face recognizers (Grabman et al., 2019). However, other authors have
shown that high recognition confidence can be misleading at individual level (Sauer et al.,
2019). This can be particularly important in other-race face identification, as it has been
shown that people are generally more overconfident when identifying other-race faces
compared to own-race faces (Dodson and Dobolyi, 2016).
In addition, findings about the reliability of recognition confidence for other-race faces are
mixed. For example, although some studies have recently found that high recognition
confidence is equally associated with the identification accuracy of own and other race
faces (Dodson and Dobolyi, 2016;Grabman et al., 2019), other studies reported that
observers not only tend to be less accurate in judging whether they will identify an other-
race face compared to an own-race face (Hourihan et al., 2012;Smith et al., 2001) but also
they tend to be have a less clear memory for other-race faces (Brigham et al.,2007;Smith
et al.,2004
). One feature of recognition confidence that might explain this disparity in the
results is its situation-specificity. In recognition confidence procedures, observers are
simply asked about their confidence about having seen a particular face. However, this
procedure is specific for the seen face and does not gather information about how good the
observer is recognizing either own- and other-race faces.
One potential alternative to recognition confidence could be self-reported measures of face
recognition (Bate et al., 2018;Bobak et al.,2019;Livingston and Shah, 2018;Shah et al.,
2015a; Ventura et al., 2018). These measures describe different daily life situations involving
face recognition abilities, so, in contrast to recognition confidence, self-reported measures
of face recognition are not situation-specific. For example, Sha and colleagues (Shah et al.,
2015a) introduced the PI20 questionnaire, a 20-statements self-reported measure of face
identification. Agreement with these statements is scaled on a five-point Likert-scale.
The authors reported strong negative correlations between the score in the PI20 and
objective measures of face identification, such as famous face recognition, the Cambridge
jJOURNAL OF CRIMINAL PSYCHOLOGY j
Face Memory Test (CFMT-Original, Duchaine and Nakayama, 2006) and the Glasgow Face
Matching Test (Shah et al.,2015b) a perceptual measure of face identification that
simulates the id-verification scenarios and that it has been used in several applied studies
with police (Robertson et al.,2016) and passport control officers (White et al., 2014).
Interestingly, the PI20 has been translated to different languages, including Portuguese
(Ventura et al., 2018), Japanese (Nakashima et al.,2020) and Mandarin (Estudillo and
Wong, 2021). In general, studies using the PI20 have concluded that adults have moderate
to strong insights into their face recognition abilities (Livingston and Shah, 2018;Ventura
et al.,2018but see Bobak et al.,2019;Estudillo and Wong, 2021).
A question that arises is whether these insights can be generalized to other-race face
recognition. At a theoretical level this has important consequence for models of
metacognition. For example, according to the well-known DunningKruger effect, unskilled
people (i.e. non-experts in a specific field) present very poor insights about their actual
performance (Dunning et al.,2003;Fakcharoenphol et al.,2015). This effect which, has
been found in different domains, including humour, logic, grammar knowledge and face
perception (Dunning et al.,2003;Pennycook et al.,2017;Zhou and Jenkins, 2020), makes a
clear prediction about our study: as people are generally experts recognizing own-race
faces, but not other-race faces (Estudillo et al., 2020;Tanaka et al.,2013), people would
have more accurate insights into their recognition abilities for own-race faces compared to
other-race faces. Thus, this study also allows us to test the DunningKruger effect from a
different domain: the recognition of own- and other-races faces. From a more applied
perspective, having insights to recognize other-race faces would be important in forensic
scenarios, such as during the identification of other-race perpetrators or during id
verification processes, as it could provide information about how reliable an observer is
identifying other-race faces. However, to date, self-reported measures of face recognition
abilities have not been adapted to other-race faces.
The present study aims to explore observers’ insight to recognize own- and other-race
faces. To achieve this, we used the PI20 to evaluate observers’ own-race face recognition
and a new developed self-reported measure for evaluating other-race face recognition (the
ORE20). Given the high reliability and construct validity of the PI20 (Shah et al., 2015a), the
ORE20 is a partial adaptation of the PI20. Participants performed objective measures for
the recognition of own- and other-race faces, the PI20 and the ORE20. We expect
participants to be better recognizing own- than other-race faces. Following previous
research with the PI20, we also expect that participants will have moderate to strong
insights into their face recognition abilities for own-race faces. This would be evident by a
negative association between the objective measure of own-race face recognition and the
PI20. Finally, if participants have insights into their recognition for other-race faces we would
expect that the ORE20 would be associated with their performance in the objective
measure of the recognition of other-race faces.
Methods
Participants
Eighty-five Chinese ethnic participants (58 females) from the University of Nottingham
Malaysia took part in this study for course credits. Observers mean age was of 21 yeas
(SD = 3). All observers reported having normal or corrected-to-normal vision. Participants
gave their informed consent and were debriefed at the end of the study. This study was
approved by the ethics committee of the University of Nottingham Malaysia.
Materials, apparatus and procedure
This study involves an objective face recognition stage and a self-reported face recognition
abilities stage. The order of these stages was counterbalanced across participants.
jJOURNAL OF CRIMINAL PSYCHOLOGY j
Testable platform was used to present stimuli and to record observers’ responses. As
English is the teaching language at the University of Nottingham Malaysia, all the
instructions and questionnaires in this study were in English.
Objective face recognition stage
In this stage, observers performed both the CFMT-Original (Duchaine and Nakayama,
2006) and the CFMT-Chinese (McKone et al.,2012) in a counter-balanced order. Both tests
follow an identical format, but they use Caucasian and Chinese faces as stimuli,
respectively. These tests are valid measures of face recognition as they require the
recognition of faces across different images, and no simple pictorial recognition (Bruce,
1982;Estudillo, 2012;Estudillo and Bindemann, 2014;Longmore et al.,2008). Participants
are required to learn and recognize different unfamiliar faces in three different stages. In the
same image stage, observers are asked to learn a target identity, which is presented in
three different orientations (i.e. mid-profile left, frontal and mid-profile right). Observers are
then presented with a three-alternative forced-choice task, whereby they are required to
select the studied target identity among two foils, with one trial per orientation. This stage is
repeated for each of six target identities, giving a total of 18 trials. In the novel images
stage, observers are presented with the same six target identities all in once for 20 s. Then,
they have to identify a novel instance of the target face among two distractors. This stage
contains a total of 30 trials. The last stage is identical to the previous stage, but faces at test
are presented with visual noise to make the task more challenging. This stage contains a
total of 24 trials, thus the maximum score an observer can obtain is 72. Internal reliability
analysis showed alpha values of 0.86 for the CFMT-Original and 0.90 for the CFMT-Chinese.
These values are in agreement with previous research (e.g. Bowles et al., 2009;Estudillo
et al.,2020
).
Self-reported face recognition abilities stage
In this stage, observers filled in both the PI20 and the ORE20 questionnaires in a counter-
balanced order. The PI20 (Shah et al.,2015a) is a self-reported measure of face
recognition. It contains 20 items describing the experience of face recognition. In the
instructions of this test, we emphasized that these items refer to the recognition of Chinese
peers (i.e. own-race faces for our sample).
The ORE20 questionnaire was developed to conduct this study. This questionnaire is a self-
reported measure of face recognition abilities for other-race faces. The questionnaire
comprises 20 items reporting different situations describing the experience of recognizing
Caucasian faces (other-race faces for our sample). Most of the items of the ORE20 were
adapted from the PI20 (e.g. Item 2 ORE20: to recognize Caucasian people, I rely on non-
facial cues, such as voice, Item 10 PI20: Without hearing peoples voices, I struggle to
recognize them; Item 10 ORE20: My face recognition ability is similar for Caucasians and
Chinese faces, Item 2 PI20: My face recognition ability is worse than most people).
However, this adaptation is not possible for some of the PI20 items (e.g. PI20 item 11:
Anxiety about face recognition has led me to avoid certain social or professional situations;
PI20 Item 13: I am very confident in my ability to recognize myself in photographs; Item 15:
My friends and family think I have bad face recognition or bad face memory). Thus, other
items of the ORE20 were created with the aim of capturing different aspects involved in the
processing of other races faces, such as learning (e.g. Item 20 ORE20: I have to try harder
to memorise Caucasian faces than Chinese faces), recognition (e.g. Item 13 ORE20: I
struggled to recognise my Caucasian classmates or colleagues), discrimination (e.g. Item
14 ORE20: I find that most Caucasian faces look alike) and race categorization (e.g. Item 9
ORE20: I mistake familiar Caucasian people as Chinese people when they change their
hairstyle).
jJOURNAL OF CRIMINAL PSYCHOLOGY j
Observers were asked to rate their agreement with each statement of the PI20 and the
ORE20 on a five-point Likert-scale (1 = strongly agree, 5 = strongly disagree). In both
questionnaires, higher total scores reflect lower recognition abilities. However, for the sake
of simplicity, the scores of the positive items were reversed (PI20: items 1, 2, 3, 4, 5, 6, 7,
10, 11, 12, 14, 15, 16, 18, 20; ORE20: items 2, 3, 4, 5, 7, 8, 9, 11, 13, 14, 15, 16, 17, 18, 20),
so that higher scores in both questionnaires indicate better self-perceived face recognition
abilities. Internal reliability analysis revealed alpha values of 0.84 and 0.89, for the PI20 and
the ORE20, respectively.
Results
In a first part of our analysis, we compared observers’ recognition performance for own and
other-race faces. Observers were better recognizing own- compared to other-race faces
[t(84) = 7.77, p<.001, d = 0.68, CI =0.48 - 0.88] [Figure 1(a)]. Interestingly, the same
pattern was also obtained at a phenomenological level as observers reported to be worse
recognizing other-race faces as reflected by their scores in the ORE20 compared to
own-race faces as reflected by their scores in the PI20 [t(84)= 6.30, p<. 001, d = 0.65,
CI = 0.42 - 0.88] [Figure 1(b)].
In a second part of our analysis, we explored observers’ insight to recognize own- and
other-race faces. Scores in the PI20 were moderately associated with the scores in the
CFMT-Chinese [r= 0.45, p<0.001, CI = 0.26 - 0.60] [Figure 2(a)]. On the other hand,
scores in the ORE20 were only poorly associated with the scores in the CFMT-Original
[r= 0.21, p = 0.04, CI =0.00 - 0.41] [Figure 2(b)]. The differences between both correlations
were explored using Silver and colleagues’ test of non-overlapping dependent correlations
(Diedenhofen and Musch, 2015;Silver et al.,2004). This test showed that the magnitude of
the association between own-race face recognition and the PI20 was stronger that the
magnitude of the association between other-race face recognition and the ORE20 (z= 2.08,
p<0.05). The PI20 was also associated with observers’ performances in the CFMT-Original
[r= 0.37, p<0.001, CI = 0.17 - 0.54] [Figure 2(c)]. However, Hittner and colleagues’ test of
overlapping dependent correlations (Diedenhofen and Musch, 2015;Hittner et al., 2003)
Figure 1 (a) Mean scores in the CFMT-Chinese and in the CFMT-Original; (b) mean
scores in the ORE20 and PI20
jJOURNAL OF CRIMINAL PSYCHOLOGY j
showed no differences between the magnitudes of this association and the association
between the ORE20 and the CFMT-Original [z= 1.58, p = 0.11].
We also explored whether scores in the ORE20 questionnaire predict the other-race
effect magnitude. We first considered the magnitude of the other-race effect as the
difference between accuracy performance in the CFMT-Chinese and the CFMT-Original
(i.e. subtraction method). In this sense, it would be expected that higher scores in the
ORE20 would be negatively associated with the other-race effect magnitude
[Figure 2(d)]. We found that the score in the ORE20 was not associated with the other-
race effect magnitude [r=0.11, p = 0.26, CI = 0.31 - 0.10]. However, it has been
argued that the subtraction method for calculating the other-race effect might hinder the
contribution of the two components of the other-race effect, namely the own-race
advantage and other-race decrement (DeGutis et al.,2013a). On the contrary,
calculating the other-race effect with a regression method allows the isolation of these
two components by regressing other- from own-race performance to produce own-race
advantage residuals and own- from other-race to produce other-race decrement
residuals (DeGutis et al., 2013a; DeGutis et al.,2013b). Thus, once the residuals for both
components were calculated, they were correlated with the scores in the ORE20
[Figure 2(e) and 2(f)]. We found that scores in the ORE20 were not associated either with
the other-race decrement [r= 0.19, p= 0.07, CI = 0.15 - 0.39] nor with the own-race
advantage [r=0.05, p=0.61,CI= 0.26 - 0.16].
Figure 2 (a) Simple correlationbetween the own-race face recognition measure (CFMT-
Chinese) and the PI20; (b) simple correlation between the other-race face
recognition measure (CFMT-Original) and theORE20; (c) simple correlation
between the other-raceface recognition measure (CFMT-Original) and the PI20;
(d) simple correlation between the other-race face effect magnitude(CFMT-
Chinese CFMT-Original) and the ORE20; (e) simple correlation between the
other-race decrement (residuals from regressing own- from other-race
performance) and the ORE20; (f) simple correlation between the own-race
decrement (residuals from regressing other- from own-race performance) and
the ORE20
jJOURNAL OF CRIMINAL PSYCHOLOGY j
In conclusion, although our results show that observers have moderate insights into their
recognition abilities for own-race faces, their insights to recognize other-race faces are very
limited.
Discussion
In recent years, there has been a growing interest in phenomenological measures of
face recognition. The PI20 questionnaire was created for this aim and has been used
with normal and prosopagnosic population (Shah et al., 2015a). Using this instrument,
several studies have shown that human observers have moderate to good insights into
their face recognition abilities (Livingston and Shah, 2018;Shah et al., 2015b; Ventura
et al., 2018). The present study aimed to explore whether observers also have insights
into their recognition of other-race faces. For this aim, we created the ORE20, a 20-item
self-reported measure of the recognition of other-race faces (Caucasian faces in our
case).
Our results show that our observers were better recognizing own-race faces compared
other-race faces. This pattern of results replicates the ORE observed in other studies
(e.g. Chiroro and Valentine, 1995;Estudillo et al., 2020;Malpass and Kravitz, 1969;
Meissner and Brigham, 2001;Wong et al., 2020). Interestingly, the ORE observed in our
study was also evident at a phenomenological level as observers reported to be worse
recognizing other-race faces compared to own-race faces. We also found a moderate
association between our objective measure of face recognition (i.e. the CFMT-Chinese in
our case) and the PI20, replicating previous research with a Chinese ethnic population.
This association was stronger in magnitude compared to the weak association found
between our objective measure of other-race face recognition (i.e. the CFMT-Original)
and the ORE20, demonstrating that the self-reported face recognition abilities are
stronger for own-races compared to other-race faces. We also found no association
between the other-race effect magnitude (calculated by both subtractions and
regression methods) and the ORE20.
Our results are in agreement with previous studies showing that observers tend to report
that their memory is less clear when they have to recognize other-race faces compared to
own-race faces (Smith et al., 2004). In fact, observers’ metamnemonic accuracy is also
worse for other-race faces (Hourihan et al., 2012). These results, in conjunction with our own
results suggests that observers do not only have remarkable problems to recognize other-
race faces but also that their insights for the recognition of other-race faces can be highly
misleading.
The weak association found between the recognition of other-race faces (CFMT-Original)
and the ORE20 cannot be explained by the psychometric properties of these tests, as our
reliability analysis showed that both tests have a strong internal consistency. However, it is
important to note that the limited insight to recognize other-race faces contrasts with the
moderated-to-good insights observers have into their general face recognition abilities, as
shown by others (Livingston and Shah, 2018;Shah et al.,2015b; Ventura et al.,2018) and
the results of the present study. These general insights into face recognition abilities can
also explain the association found between the PI20 and the recognition of other-race faces.
Despite the large differences found in terms of accuracy, research has shown that the
recognition of own- and other-race faces share around 45% of variance ( Wan et al., 2017).
Thus, the association between the PI20 and the CFMT-Original seem to reflect these
general face recognition processes, and not the specific processes of recognizing other-
race faces. From these results, it could be argued that, compared to the ORE20, the PI20 is
a better proxy of other-race face recognition and, therefore, could be potentially used as an
index of the recognition of other-race faces. However, one should be cautious about this
conclusion, as the magnitude of the association found between other-race face recognition
jJOURNAL OF CRIMINAL PSYCHOLOGY j
and the PI20 was not different to that found between other-race face recognition and the
ORE20.
Our results support theories of metacognition that claims that insights about performance
are modulated by experience (Dunning et al., 2003). Specifically, as human observers
tend to have extensive experience with own-race faces, they have considerable number
of opportunities to evaluate their face recognition skills for own-race face. The lack of
experience with other-race faces, thus, limits their evaluation of face recognition skills
with other-race faces. This lack of experience also explains why our observers are still
aware that, in comparison to the recognition of own-race faces, they are worse
recognizing other-race faces. Lack of experience with a particular racial group produces
problems to discriminate faces from that group. This is such that people without
experience with a racial group tend to report that faces from that group tend to look alike
(Meissner and Brigham, 2001). Thus, observers would know that they will not be able to
recognize other-race faces. For this reason, they will rate worse their recognition abilities
for other-race faces compared to own-race faces. Although given the differences
between the PI20 and the ORE20 questionnaires, we are cautious about this conclusion,
converging evidence from expertise research seems to support this hypothesis, as
experts have greater metacognitive awareness of their expertise ability compared to
novices (Dunning et al., 2003;Fakcharoenphol et al.,2015;Persky and Robinson, 2017).
To determine whether and how experience with other-race faces affects the insights into
other-race face recognition, future research should include measures of contact with
other races (Zhao et al.,2014).
In conclusion, our results show that although observers have moderate insights into their
recognition abilities for own-race faces, their insights for other-race face recognition abilities
are limited. These results have important consequences in applied face recognition
settings. For example, in forensic settings, such as line-up identification parades, it might
be necessary to assess an eyewitness ability to recognize faces. In these scenarios, self-
reported measures of other-race recognition could produce misleading results. For this
reason, when evaluating eyewitness’ accuracy identifying other-race faces, objective
measures should be used. Although, self-reported measures of own-race faces could
provide some indication of observers’ actual face recognition abilities, at the best, these
measures should be used as a complement of objective face recognition measures.
References
Bate, S., Frowd, C., Bennetts, R., Hasshim, N., Murray, E., Bobak, A.K., Wills, H. and Richards, S. (2018),
“Applied screening tests for the detection of superior face recognition”, Cognitive Research: Principles
and Implications, Vol. 3 No. 1, p. 22, doi: 10.1186/s41235-018-0116-5.
Bobak, A.K., Mileva, V.R. and Hancock, P.J. (2019), “Facing the facts: naive participants have only
moderate insight into their face recognition and face perception abilities”, Quarterly Journal of
Experimental Psychology, Vol. 72 No. 4,pp. 872-881, doi: 10.1177/1747021818776145.
Bowles, D.C., McKone, E., Dawel, A., Duchaine, B., Palermo, R., Schmalzl, L., Rivolta, D., Wilson, C.E.
and Yovel, G. (2009), “Diagnosing prosopagnosia: effects of ageing, sex, and participantstimulus ethnic
match on the Cambridge face memory test and Cambridge face perception test”, Cognitive
Neuropsychology, Vol. 26 No. 5, pp. 423-455.
Brigham, J.C., Bennett, L.B., Meissner, C.A. and Mitchell, T.L. (2007), “The influence of race on eyewitness
memory”, in Lindsay, R.C.L., Ross, D.F., Read, J.D. and Toglia, M.P. (Eds), The Handbook of Eyewitness
Psychology,Vol.2.MemoryforPeople, Lawrence Erlbaum Associates Publishers, Mahwah, NJ, pp. 257-281.
Bruce, V. (1982), “Changing faces: visual and non-visual coding processes in face recognition”, British
Journal of Psychology, Vol. 73 No. 1, pp. 105-116.
Chiroro, P. and Valentine, T. (1995), “An investigation of the contact hypothesis of the own-race bias
in face recognition”, The Quarterly Journal of Experimental Psychology Section A,Vol.48No.4,
pp. 879-894, doi: 10.1080/14640749508401421.
jJOURNAL OF CRIMINAL PSYCHOLOGY j
DeGutis, J., Mercado, R.J., Wilmer, J., Rosenblatt, A., Meissner, C., Brigham, J., ...Curran, T. (2013a),
“Individual differences in holistic processing predict the own-race advantage in recognition memory”,
PLoS ONE, Vol. 8 No. 4, p. e58253, doi: 10.1371/journal.pone.0058253.
DeGutis, J., Wilmer, J., Mercado, R.J. and Cohan, S. (2013b), “Using regression to measure holistic face
processing reveals a strong link with face recognition ability”, Cognition, Vol. 126 No. 1, pp. 87-100, doi:
10.1016/j.cognition.2012.09.004.
Diedenhofen, B. and Musch, J. (2015), “Cocor: a comprehensive solution for the statistical comparison of
correlations”, Plos One, Vol. 10 No. 4, p.e0121945, doi: 10.1371/journal.pone.0121945.
Dodson, C.S. and Dobolyi, D.G. (2016), “Confidence and eyewitness identifications: the cross-race
effect, decision time and accuracy”, Applied Cognitive Psychology, Vol. 30 No. 1, pp. 113-125, doi:
10.1002/acp.3178.
Duchaine, B. and Nakayama, K. (2006), “The Cambridge face memory test: results for neurologically intact
individuals and an investigation of its validity using inverted face stimuli and prosopagnosic participants”,
Neuropsychologia, Vol. 44 No. 4, pp. 576-585, doi: 10.1016/J.NEUROPSYCHOLOGIA.2005.07.001.
Dunning, D., Johnson, K., Ehrlinger, J. and Kruger, J. (2003), “Why people fail to recognize their own
incompetence”, Current Directions in Psychological Science, Vol. 12No. 3, pp. 83-87, doi: 10.1111/1467-
8721.01235.
Estudillo, A.J. (2012), “Facial memory: the role of the pre-existing knowledge in face processing and
recognition”, Europe’s Journal of Psychology, Vol. 8 No. 2, pp. 231-244, doi: 10.5964/ejop.v8i2.455.
Estudillo, A.J. and Bindemann, M. (2014), “Generalization across view in face memory and face
matching”, i-Perception, Vol. 5 No. 7, pp. 589-601, doi: 10.1068/i0669.
Estudillo, A.J. and Wong, H.K. (2021), “Associations between self-reported and objective face
recognition abilities are only evident in above-and below-average recognisers”, PeerJ, Vol. 9, p. e10629.
Estudillo, A.J., Lee, J.K.W., Mennie, N. and Burns, E. (2020), “No evidence of other-race effect for
Chinese faces in Malaysian non-Chinese population”, Applied Cognitive Psychology, Vol. 34 No. 1,
pp. 270-276, doi: 10.1002/acp.3609.
Evans, J.R., Marcon, J.L. and Meissner, C.A. (2009), “Cross-racial lineup identification: assessing the
potential benefits of context reinstatement”, Psychology, Crime & Law, Vol. 15 No. 1, pp. 19-28,
doi: 10.1080/10683160802047030.
Fakcharoenphol, W., Morphew, J.W. and Mestre, J.P. (2015), “Judgments of physics problem difficulty
among experts and novices”, Physical Review Special Topics - Physics Education Research, Vol. 11
No. 2, p. 020128, doi: 10.1103/PhysRevSTPER.11.020128.
Grabman, J.H., Dobolyi, D.G., Berelovich, N.L. and Dodson, C.S. (2019), “Predicting high confidence
errors in eyewitness memory: the role of face recognition ability, decision-time, and justifications”, Journal
of Applied Research in Memory and Cognition, Vol. 8 No. 2, pp. 233-243, doi: 10.1016/J.
JARMAC.2019.02.002.
Hittner, J.B., May, K. and Silver, N.C. (2003), “A Monte Carlo evaluation of tests forcomparing dependent
correlations”, The Journal of General Psychology, Vol. 130 No. 2, pp. 149-168, doi: 10.1080/
00221300309601282.
Hourihan, K.L., Benjamin, A.S. and Liu, X. (2012), “A cross-race effect in metamemory: predictions of
face recognition are more accurate for members of our own race”, Journal of Applied Research in
Memory and Cognition, Vol. 1 No. 3, pp. 158-162, doi: 10.1016/J.JARMAC.2012.06.004.
Kokje, E., Bindemann, M. and Megreya, A.M. (2018), “Cross-race correlations in the abilities to match
unfamiliar faces”, Acta Psychologica, Vol. 185, pp. 13-21, doi: 10.1016/J.ACTPSY.2018.01.006.
Livingston, L.A. and Shah, P. (2018), “Article commentary: people with and without prosopagnosia have
insight into their face recognition ability”, Quarterly Journal of Experimental Psychology, Vol. 71 No. 5,
pp. 1260-1262, doi: 10.1080/17470218.2017.1310911.
Longmore, C. A., Liu, C.H. and Young, A.W. (2008), “Learning faces from photographs”, Journal of
Experimental Psychology: Human Perception and Performance, Vol. 34 No. 1, pp. 77-100, doi: 10.1037/
0096-1523.34.1.77.
McKone, E., Stokes, S., Liu, J., Cohan, S., Fiorentini, C., Pidcock, M., Yovel, G., Broughton, M.and Pelleg,
M. (2012), “A robust method of measuring other-race and other-ethnicity effects: the Cambridge face
memory test format”, PLoS ONE, Vol. 7 No. 10, p. e47956, doi: 10.1371/journal.pone.0047956.
jJOURNAL OF CRIMINAL PSYCHOLOGY j
Malpass, R.S. andKravitz, J. (1969), “Recognition for faces of own and other race”, Journal of Personality
and Social Psychology, Vol. 13 No. 4, pp. 330-334, doi: 10.1037/h0028434.
Megreya, A.M., White, D. and Burton, A.M. (2011), “The other-race effect does not rely on memory:
evidence from a matching task”, Quarterly Journal of Experimental Psychology,Vol.64No.8,
pp. 1473-1483, doi: 10.1080/17470218.2011.575228.
Meissner, C.A. and Brigham, J.C. (2001), “Thirty years of investigating the own-race bias in memory for
faces: a meta-analytic review”, Psychology, Public Policy, and Law, Vol. 7 No. 1, pp. 3-35, doi: 10.1037/
1076-8971.7.1.3.
Nakashima, S.F., Ukezono, M., Sudo, R., Nunoi, M., Kitagami, S., Okubo, M., Toriyama, R., Morimoto, Y.
and Takano, Y. (2020), “Development of a Japaneseversion of the 20-item prosopagnosia index (PI20-J)
and examination of its reliability and validity”, The Japanese Journal of Psychology, Vol. 90 No. 6,
pp. 603-613, doi: 10.4992/jjpsy.90.18235.
Pennycook, G., Ross, R.M., Koehler, D.J. and Fugelsang, J.A. (2017), “DunningKruger effects in
reasoning: theoretical implications of the failure to recognize incompetence”, Psychonomic Bulletin &
Review, Vol. 24 No. 6,pp. 1774-1784, doi: 10.3758/s13423-017-1242-7.
Persky, A.M. and Robinson, J.D. (2017), “Moving from novice to expertise and its implications for
instruction”, American Journal of Pharmaceutical Education, Vol. 81 No. 9, p. 6065, doi: 10.5688/
ajpe6065.
Potter, R. and Brewer, N. (1999), “Perceptions of witness behaviour-accuracy relationships held by
police, lawyers and mock-jurors”, Psychiatry, Psychology and Law, Vol. 6 No. 1, pp. 97-103, doi: 10.1080/
13218719909524952.
Robertson, D.J., Noyes, E., Dowsett, A., Jenkins, R., Burton, A.M. and Burton, M. (2016), “Face
recognition by metropolitan police super-recognisers correspondence to”, PLoS One,Vol.11No.2,
p. e0150036, doi: 10.1371/journal.pone.0150036.
Sauer, J.D., Palmer, M.A. and Brewer, N. (2019), “Pitfalls in using eyewitness confidence to diagnose the
accuracy of an individual identification decision”, Psychology, Public Policy, and Law,Vol.25No.3,
pp. 147-165, doi: 10.1037/law0000203.
Shah, P., Gaule, A., Sowden, S., Bird, G. and Cook, R. (2015a), “The 20-item prosopagnosia index (PI20)
: a self-report instrument for identifying developmental prosopagnosia”, Royal Society Open Science,
Vol. 2 No. 6, pp. 1-11,doi: 10.1098/rsos.140343.
Shah, P., Sowden, S., Gaule, A., Catmur, C. and Bird, G. (2015b), “The 20 item prosopagnosia index
(PI20): relationship with the Glasgow face-matching test”, Royal Society Open Science, Vol. 2 No. 11,
pp. 0-5, doi: 10.1098/rsos.150305.
Silver, C., Hittner, J. and May, K. (2004), “Testing dependent correlations with nonoverlapping variables:
a Monte Carlo simulation”, The Journal of Experimental Education, Vol. 73 No. 1, pp. 53-69.
Smith, S.M., Lindsay, R.C.L., Pryke, S. and Dysart, J.E. (2001), “Postdictors of eyewitness errors: can
false identifications be diagnosed in the cross-race situation?”, Psychology, Public Policy, and Law,
Vol. 7 No. 1, pp. 153-169, doi: 10.1037/1076-8971.7.1.153.
Smith, S.M., Stinson, V. and Prosser, M.A. (2004), “Do they all look alike? An exploration of decision-
making strategies in cross-race facial identifications”, Canadian Journal of Behavioural Science/Revue
Canadienne Des Sciences du Comportement, Vol. 36 No. 2, pp. 146-154, doi: 10.1037/h0087225.
Tanaka, J.W., Heptonstall, B. and Hagen, S. (2013), “Perceptual expertise and the plasticity of
other-race face recognition”, Visual Cognition, Vol. 21 Nos 9/10, pp. 1183-1201, doi: 10.1080/
13506285.2013.826315.
Ventura, P., Livingston, L.A. and Shah, P. (2018), “Adults have moderate-to-good insight into their face
recognition ability: further validation of the 20-item prosopagnosia index in a Portuguese sample”,
Quarterly Journal of Experimental Psychology, Vol. 71 No. 12, pp. 2677-2679, doi: 10.1177/
1747021818765652.
Wan, L., Crookes, K., Dawel, A., Pidcock, M., Hall, A. and McKone, E. (2017), “Face-blind for other-race
faces: individual differences inother-race recognition impairments”, Journal of Experimental Psychology:
General, Vol. 146 No. 1, pp. 102-122, doi: 10.1037/xge0000249.
White, D., Kemp, R.I., Jenkins, R., Matheson, M. and Burton, A.M. (2014), “Passport officers’ errors in face
matching”, PLoS ONE, Vol. 9 No. 8, p. e103510, doi: 10.1371/journal.pone.0103510.
jJOURNAL OF CRIMINAL PSYCHOLOGY j
Wixted, J.T. and Wells, G.L. (2017), “The relationship between eyewitness confidence and identification
accuracy: a new synthesis”, Psychological Science in the Public Interest, Vol. 18 No. 1, pp. 10-65,
doi: 10.1177/1529100616686966.
Wong, H.K., Stephen, I.D. and Keeble, D.R.T. (2020), “The own-race bias for face recognition in a
multiracial society”, Frontiers in Psychology, Vol. 11, p. 208, doi: 10.3389/fpsyg.2020.00208.
Zhao, M., Hayward, W.G. and Bu
¨lthoff, I. (2014), “Holistic processing, contact, and the other-race effect
in face recognition”,Vision Research, Vol. 105, pp. 61-69, doi: 10.1016/j.visres.2014.09.006.
Zhou, X. and Jenkins, R. (2020), “DunningKruger effects in face perception”, Cognition, Vol. 203,
p. 104345, doi: 10.1016/j.cognition.2020.104345.
Corresponding author
Alejandro J. Estudillo can be contacted at: aestudillo@bournemouth.ac.uk
For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com
jJOURNAL OF CRIMINAL PSYCHOLOGY j
... This allowed us to isolate the contribution of holistic and featural processing during the recognition stage to FRA. In addition, to measure individual differences in face recognition, observers performed the Cambridge Face Memory Test-Chinese (CFMT-Chi) 35 , a highly reliable and valid measure of individual differences in face recognition skills 36 . ...
... The maximum achievable score (e.g., sum of correct responses) for the CFMT-Chi is 72, in which our current sample had a mean score of 57.98 (SD = 8.93) in Experiment 1 and 58.28 (SD = 8.53) in Experiment 2. As revealed by a two-tailed independent-samples t-test, the mean CFMT-Chi scores for both experiments were not significantly different from each other, t(171) = − 0.23, p = 0.820, η p 2 = − 0.035. This shows that our participants' FRA are largely similar between the two experiments, as well as with those of previous studies [35][36][37][38][39] . Mean accuracy scores of the RMT were calculated separately for each of the two viewing conditions: "whole" and "aperture" (Fig. 2). ...
Article
Full-text available
While it is generally accepted that holistic processing facilitates face recognition, recent studies suggest that poor recognition might also arise from imprecise perception of local features in the face. This study aimed to examine to what extent holistic and featural processing relates to individual differences in face recognition ability (FRA), during face learning (Experiment 1) and face recognition (Experiment 2). Participants performed two tasks: (1) The “Cambridge Face Memory Test-Chinese” which measured participants’ FRAs, and (2) an “old/new recognition memory test” encompassing whole faces (preserving holistic and featural processing) and faces revealed through a dynamic aperture (impairing holistic processing but preserving featural processing). Our results showed that participants recognised faces more accurately in conditions when holistic information was preserved, than when it is impaired. We also show that the better use of holistic processing during face learning and face recognition was associated with better FRAs. However, enhanced featural processing during recognition, but not during learning, was related to better FRAs. Together, our findings demonstrate that good face recognition depends on distinct roles played by holistic and featural processing at different stages of face recognition.
... First, some researchers argue that self-report questionnaires like the PI20 do not accurately depict face recognition abilities (Matsuyoshi & Watanabe, 2021;Palermo et al., 2017). This is because the associations between such questionnaires only weakly-to-moderately correlate with performance on cognitive tasks such as the CFMT Estudillo, 2021;Estudillo & Wong, 2021;Matsuyoshi & Watanabe, 2021;Shah et al., 2015;Palermo et al., 2017;Tsantani et al., 2021;Ventura et al., 2018). Given that experimental tests provide objective data based on experiences designed to imitate those found in the real world (e.g., viewing a face and having to identify it), researchers assume that they must engage actual processes used in day-to-day life. ...
... These discussions are common in other neurodevelopmental disorders, such as autism(Abu-Akel et al. 2019;Chown & Leatherland, 2020; Happé & Firth, 2020, 2021Elton et al., 2016;Kim et al. 2019;Tang et al. 2020) and dyslexia(Cilibrasi & Tsimpli, 2020;Wagner et al. 2020). Both conditions have garnered a great deal more research into them than prosopagnosia (Google Scholar's number of papers with term in title: 'autism', 213000; 'dyslexia', 24800; 'developmental prosopagnosia', 308), and yet, authors are still uncertain which hypothesis best fits the data. ...
Article
Full-text available
Developmental prosopagnosia is characterised by severe, lifelong difficulties when recognising facial identity. Unfortunately, the most common diagnostic assessment (Cambridge Face Memory Test) misses 50-65% of individuals who believe that they have this condition. This results in such excluded cases’ absence from scientific knowledge, effect sizes of impairment potentially overestimated, treatment efficacy underrated, and may elicit in them a negative experience of research. To estimate their symptomology and group level impairments in face processing, we recruited a large cohort who believes that they have prosopagnosia. Matching prior reports, 56% did not meet criteria on the Cambridge Face Memory Test. However, the severity of their prosopagnosia symptoms and holistic perception deficits were comparable to those who did meet criteria. Excluded cases also exhibited face perception and memory impairments that were roughly one standard deviation below neurotypical norms, indicating the presence of objective problems. As the prosopagnosia index correctly classified virtually every case, we propose it should be the primary method for providing a diagnosis, prior to subtype categorisation. We present researchers with a plan on how they can analyse these excluded prosopagnosia cases in their future work without negatively impacting their traditional findings. We anticipate such inclusion will enhance scientific knowledge, more accurately estimate effect sizes of impairments and treatments, and identify commonalities and distinctions between these different forms of prosopagnosia. Owing to their atypicalities in visual perception, we recommend that the prosopagnosia index should be used to screen out potential prosopagnosia cases from broader vision research.
... Third, most studies on the CRE have used Black and White faces, with limited attention given to other races, such as Asian faces (cf. Estudillo, 2021;Gross, 2009;Smith et al., 2001), which we examined. ...
Preprint
Full-text available
The Cross-Race Effect (CRE) is a reliable and robust phenomenon, whereby individuals better recognize faces that belong to their race compared to another race. Our goal was to produce items for a self-report Inventory (i.e., CRE-I) that brings together known predictors of the CRE to improve the postdiction of cross-race eyewitness accuracy.We expected a CRE for White and Asian participants. We anticipated that developed CRE-I subscales would correlate positively with extant (some modified) scales and predict accuracy. Participants completed four trials (two White targets and two Asian targets). For each trial, they watched a mock crime video, performed a distractor task, made a sequential lineup decision (target-present or target-absent), and indicated confidence in their lineup decision. After all trials, participants completed the potential items for the CRE-I.We replicated prior findings of a CRE for White participants but did not find a CRE for Asian participants. Exploratory factor analysis produced internally reliable scales for the CRE-I to be used with White eyewitnesses: general face recognition ability, race-specific face recognition ability, racial attitudes, quantity of contact, quality of contact, motivated individuation, and cognitive disregard. Responses to several scales predicted identification accuracy. In particular, three CRE-I scales predicted identification accuracy beyond the predictiveness of confidence: race-specific face recognition ability, racial attitudes towards White people, and motivated individuation of White people. Variables suggested separately by the perceptual expertise hypothesis and the social cognitive hypothesis predicted identification accuracy, providing support for integrative models of the CRE. The CRE-I contributes to the CRE literature both in terms of theory—by showing which factors among many may best relate to recognition—and practice—by improving evaluations of eyewitness reliability.
... Despite the simultaneous presentation of both faces, face masks also disrupted performance in face matching. Interestingly, using the Cambridge Face Memory-a highly reliable and valid measure of face memory [21,[47][48][49][50][51]]-Freud and colleagues [42] showed that the inversion effect is reduced by the presence of face masks. If it is assumed that inverting a face disrupts holistic processing [3,11,12] these results suggest that face masks also impair holistic processing. ...
Article
Full-text available
Although it is generally accepted that face recognition relies on holistic processing, it has been suggested that the simultaneous face matching task may depend on a more analytical or featural processing approach. However, empirical evidence supporting this claim is limited. In two experiments, we further explored the role of holistic and featural processing on simultaneous face matching by manipulating holistic processing through inversion and presenting faces with or without face masks. The results from Experiment 1 revealed that both inversion and face masks impaired matching performance. However, while the inversion effect was evident in both full-view and masked faces, the mask effect was only found in upright, but not inverted, faces. These results were replicated in Experiment 2 but, the inversion and mask effects were stronger in delayed face matching than in simultaneous face matching. Our findings suggest that simultaneous face matching relies on holistic processing, but to a smaller extent compared to higher memory-demanding identification tasks.
... Sixty Malaysian Chinese (38 females) participants were recruited. To address the potential influence of ORE in face recognition (Estudillo, 2021;Estudillo et al., 2020;Hayward et al., 2008;Wong et al., 2020), only Malaysian Chinese participants were recruited as the stimuli presented in the experiment were created using only Chinese faces. Past work has also indicated that hormone levels which fluctuate among females due to the menstrual cycle could affect cortical excitability (Smith et al., 2002). ...
Article
Full-text available
The functional role of the occipital face area (OFA) and the fusiform face area (FFA) in face recognition is inconclusive to date. While some research has shown that the OFA and FFA are involved in early (i.e., featural processing) and late (i.e., holistic processing) stages of face recognition respectively, other research suggests that both regions are involved in both early and late stages of face recognition. Thus, the current study aims to further examine the role of the OFA and the FFA using multifocal transcranial direct current stimulation (tDCS). In Experiment 1, we used computer-generated faces. Thirty-five participants completed whole face and facial features (i.e., eyes, nose, mouth) recognition tasks after OFA and FFA stimulation in a within-subject design. No difference was found in recognition performance after either OFA or FFA stimulation. In Experiment 2 with 60 participants, we used real faces, provided stimulation following a between-subjects design and included a sham control group. Results showed that FFA stimulation led to enhanced efficiency of facial features recognition. Additionally, no effect of OFA stimulation was found for either facial feature or whole face recognition. These results suggest the involvement of FFA in the recognition of facial features.
... In addition to detecting face processing difficulties in the neurotypical population, it can also identify people with developmental prosopagnosia, which is a characterized by an inability in recognizing faces (Bate et al., 2014;Burns, Bennetts, et al., 2017;Burns et al., 2014Burns & Bukach, 2021Childs et al., 2021;Corrow et al., 2016;Maw et al., 2023;Wilcockson et al., 2020). The PI20 has been shown to correlate with multiple measures of face memory, e.g., Cambridge Face Memory Task and Famous Faces Tests (Estudillo, 2021;Shah et al., 2015) supporting its validity on investigating face memory abilities, specifically on how people rate their own face recognition ability. It has recently been shown that the PI20 may be more effective than cognitive task-based approaches for detecting severe levels of face memory problems . ...
Article
Full-text available
Background: Risky alcohol use is related to a variety of cognitive impairments, including memory and visuo-perceptual difficulties. Remarkably, no prior work has assessed whether usage of alcohol can predict difficulties perceiving facial identity. Objectives: Therefore, this study aimed to investigate whether riskier alcohol consumption predicted impairments in face perception and self-reported difficulties in face recognition. Results: Participants (N = 239, male = 77) were over 18 years old and had normal or corrected-to-normal vision. Alcohol use was assessed using the Alcohol Use Disorder Identification Test (AUDIT), while face recognition difficulties were determined by the 20-item Prosopagnosia Index questionnaire (PI20). A subsample of participants (N = 126, male = 51) completed the Cambridge Face Perception task (CFPT) to assess their face perception ability. Multiple linear regressions showed significant models of prediction on both face perception and face recognition when considering AUDIT score and age as predictors. Conclusion: This study suggested, for the first time, that risky alcohol use predicts both poorer visuo-perceptual processing for faces and self-reported difficulties in face recognition.
... Although the estimated prevalence of developmental prosopagnosia in the general population is around 2.5% (Bowles et al., 2009;Kennerknecht et al., 2006Kennerknecht et al., , 2008, many cases remain undiagnosed (Duchaine, 2000). Given the limited insights that people have into their own face recognition skills (Bate & Dudfield, 2019;Bobak et al., 2019;Estudillo, 2021;Palermo et al., 2017), objective measures of face identification are crucial for the study of individual differences in face recognition skills and the diagnosis of prosopagnosia. ...
Article
Full-text available
The Cambridge Face Memory Test (CFMT) is one of the most important measures of individual differences in face recognition and for the diagnosis of prosopagnosia. Having two different CFMT versions using a different set of faces seems to improve the reliability of the evaluation. However, at the present time, there is only one Asian version of the test. In this study, we present the Cambridge Face Memory Test - Chinese Malaysian (CFMT-MY), a novel Asian CFMT using Chinese Malaysian faces. In Experiment 1, Chinese Malaysian participants (N = 134) completed two versions of the Asian CFMT and one object recognition test. The CFMT-MY showed a normal distribution, high internal reliability, high consistency and presented convergent and divergent validity. Additionally, in contrast to the original Asian CFMT, the CFMT-MY showed an increasing level of difficulties across stages. In Experiment 2, Caucasian participants (N = 135) completed the two versions of the Asian CFMT and the original Caucasian CFMT. Results showed that the CFMT-MY exhibited the other-race effect. Overall, the CFMT-MY seems to be suitable for the diagnosis of face recognition difficulties and could be used as a measure of face recognition ability by researchers who wish to examine face-related research questions such as individual differences or the other-race effect.
... Indeed, research has shown that unfamiliar face matching skills present substantial individual differences across observers, with some individuals performing at chance levels while others performing at ceiling levels (Bruce et al., 2018;Burton et al., 2010;Estudillo & Bindemann, 2014;McCaffery et al., 2018). Thus, this account highlights the importance of using objective face identification tasks during personnel selection for those applied settings whereby the identification of others is demanded (Bobak et al., 2016;Estudillo, 2021;Fysh et al., 2020;Ramon et al., 2019;Robertson et al., 2016). In addition, according to the data limit account, a large variance of errors in face matching can be explained by the properties of the face stimuli (Estudillo & Bindemann, 2014;Fysh & Bindemann, 2017). ...
Article
Full-text available
Although the positive effects of congruency between stimuli are well replicated in face memory paradigms, mixed findings have been found in face matching. Due to the current COVID-19 pandemic, face masks are now very common during daily life outdoor activities. Thus, the present study aims to further explore congruency effects in matching faces partially occluded by surgical masks. Observers performed a face matching task consisting of pairs of faces presented in full view (i.e., full-view condition), pairs of faces in which only one of the faces had a mask (i.e., one-mask condition), and pairs of faces in which both faces had a mask (i.e., two-mask condition). Although face masks disrupted performance in identity match and identity mismatch trials, in match trials, we found better performance in the two-mask condition compared to the one-mask condition. This finding highlights the importance of congruency between stimuli on face matching when telling faces together.
Preprint
Full-text available
This article challenges the wide-spread practice of using Hu and Bentler's (1999) cut-off values (CFI > .95, TLI > .95, RMSEA < .06, SRMR < .08) as the sole indicator of good test fit. Two underlying assumptions are questioned: (1) that these criteria apply universally and (2) that very high CFI/TLI and low RMSEA/SRMR represent optimal fit. An exclusive focus on Hu and Bentler's criteria favors top-down test construction with narrow, pre-defined scales, potentially neglecting bottom-up approaches that identify correlations reflecting real-world complexities. This article proposes an alternative approach: compiling practice-based distributions of fit indices from existing tests. An analysis of 161 datasets revealed distinct high-probability regions for TLI, RMSEA, and SRMR, largely outside the Hu and Bentler thresholds and dependent on the statistical technique used for analysis. Based on the distributions, indicators are developed to differentiate between unifactorial, multifactorial, and hierarchical scales. The indicators are intended to complement, not replace, statistical model comparison by providing a practice-oriented perspective.
Preprint
Full-text available
In the forensic face matching task, observers are presented with two unfamiliar faces and must determine whether they depict the same identity. Due to the Covid-19 pandemic, some governmental authorities require the use of face masks in public spaces. However, recent research has shown that face masks impair face identification. The present study explores the effect of face masks on forensic face matching using an individual differences approach. Compared to a full-view condition, performance decreased when a face mask was superimposed on one face (Experiment 1) and both faces (Experiment 2) of a pair. Although a positive correlation between the full-view and the mask conditions was found, high proficiency in the full-view condition did not always generalize to the mask condition. Additionally, the mask generally has a more negative impact on those participants with better performance in the full-view condition. The theoretical and practical implications of these findings are discussed.
Article
Full-text available
The 20-Item Prosopagnosia Items (PI-20) was recently introduced as a self-report measure of face recognition abilities and as an instrument to help the diagnosis of prosopagnosia. In general, studies using this questionnaire have shown that observers have moderate to strong insights into their face recognition abilities. However, it remains unknown whether these insights are equivalent for the whole range of face recognition abilities. The present study investigates this issue using the Mandarin version of the PI-20 and the Cambridge Face Memory Test Chinese (CFMT-Chinese). Our results showed a moderate negative association between the PI-20 and the CFMT-Chinese. However, this association was driven by people with low and high face recognition ability, but absent in people within the typical range of face recognition performance. The implications of these results for the study of individual differences and the diagnosis of prosopagnosia are discussed.
Article
Full-text available
The own-race bias (ORB) is a reliable phenomenon across cultural and racial groups where unfamiliar faces from other races are usually remembered more poorly than own-race faces (Meissner and Brigham, 2001). By adopting a yes–no recognition paradigm, we found that ORB was pronounced across race groups (Malaysian–Malay, Malaysian–Chinese, Malaysian–Indian, and Western–Caucasian) when faces were presented with only internal features (Experiment 1), implying that growing up in a profoundly multiracial society does not necessarily eliminate ORB. Using a procedure identical to Experiment 1, we observed a significantly greater increment in recognition performance for other-race faces than for own-race faces when the external features (e.g. facial contour and hairline) were presented along with the internal features (Experiment 2)—this abolished ORB. Contrary to assumptions based on the contact hypothesis, participants’ self-reported amount of interracial contact on a social contact questionnaire did not significantly predict the magnitude of ORB. Overall, our findings suggest that the level of exposure to other-race faces accounts for only a small part of ORB. In addition, the present results also support the notion that different neural mechanisms may be involved in processing own- and other-race faces, with internal features of own-race faces being processed more effectively, whereas external features dominate representations of other-race faces.
Article
Full-text available
The 20-item prosopagnosia index (PI20) was developed for assessing congenital prosopagnosia, which is characterized by severe facial recognition deficits in the absence of any obvious neurological deficit. We aimed to develop a Japanese version of the PI20 (PI20-J) scale and evaluate its validity and reliability. In study 1, we confirmed the internal consistency, test-retest reliability, and concurrent validity of the scale. In study 2, we examined the relationships between PI20-J score and facial recognition performance and found a moderate correlation between them. In study 3, we examined whether the PI20-J score is related specifically with facial recognition performance, or with general object recognition performance. We found that participants with a high PI20-J score showed weaker facial recognition performance than those with a low PI20-J score. In contrast, the object recognition performance did not depend on the score. Our results suggest that the PI20-J score is related specifically with facial recognition performance. We conclude that PI20-J is highly reliable and valid as a self-reported measure for congenital prosopagnosic traits.
Article
Full-text available
Recently, a number of authors have made strong claims about the likely very high accuracy of identifications made with very high levels of confidence when identification testing conditions are pristine. We argue that although these strong claims about the confidence–accuracy relation are justifiable at the aggregate level, they may be misleading when attempting to evaluate the accuracy of an individual identification. First, we consider the recent evolution of conclusions drawn about the confidence–accuracy relationship, and the implications of these conclusions for the utility of confidence for evaluating individual identifications. Next, we highlight factors that may undermine the generalizability of conclusions at the aggregate level to individual cases. Finally, we present reanalyses of published data demonstrating conditions where conclusions based on aggregate data would be misleading for practitioners evaluating an individual identification. We maintain that, when appropriately collected, confidence can be a useful guide when assessing the reliability of identifications. However, we argue that when police and triers of fact attempt to evaluate the likely accuracy of an individual identification decision it will often be impossible to know if one of the key prerequisites for assessing whether a high confidence identification indicates an accurate identification—namely, a fair lineup—has been met.
Article
Full-text available
Increasing research shows that high eyewitness confidence at the time of an initial identification is a strong predictor of accuracy (Wixted & Wells, 2017). However, as with all forms of criminal evidence, this relationship is imperfect. This study addresses whether there are variables that systematically influence the rate of high confidence misidentifications. Notably, this is the first study to document the influence of face recognition ability on the confidence-accuracy relationship. Participants viewed photos of individuals of their same race or a different race, and performed a lineup recognition test after either a 5-minute (n = 277) or 1-day (n = 292) delay. High confidence identification errors were more likely when a) individuals are worse face recognizers, b) decision-times are slow, and c) responses are justified with references to familiarity (e.g., “He looks familiar).
Article
Full-text available
In recent years there has been growing interest in the identification of people with superior face recognition skills, for both theoretical and applied investigations. These individuals have mostly been identified via their performance on a single attempt at a tightly controlled test of face memory—the long form of the Cambridge Face Memory Test (CFMT+). The consistency of their skills over a range of tests, particularly those replicating more applied policing scenarios, has yet to be examined systematically. The current investigation screened 200 people who believed they have superior face recognition skills, using the CFMT+ and three new, more applied tests (measuring face memory, face matching and composite-face identification in a crowd). Of the sample, 59.5% showed at least some consistency in superior face recognition performance, although only five individuals outperformed controls on overall indices of target-present and target-absent trials. Only one participant outperformed controls on the Crowds test, suggesting that some applied face recognition tasks require very specific skills. In conclusion, future screening protocols need to be suitably thorough to test for consistency in performance, and to allow different types of superior performer to be detected from the outset. Screening for optimal performers may sometimes need to directly replicate the task in question, taking into account target-present and target-absent performance. Self-selection alone is not a reliable means of identifying those at the top end of the face recognition spectrum. Electronic supplementary material The online version of this article (10.1186/s41235-018-0116-5) contains supplementary material, which is available to authorized users.
Article
The Dunning–Kruger Effect refers to a common failure of metacognitive insight in which people who are incompetent in a given domain are unaware of their incompetence. This effect has been found in a wide range of tasks, raising the question of whether there is any ‘special’ domain in which it is not found. One plausible candidate is face perception, which has sometimes been thought to be ‘special’. To test this possibility, we assessed participants' insight into their own face perception abilities (self-estimates) and those of other people (peer estimates). We found classic Dunning–Kruger Effects in matching tasks for unfamiliar identity, familiar identity, gaze direction, and emotional expression. Low performers overestimated themselves, and high performers underestimated themselves. Interestingly, participants' self-estimates were more stable across tasks than their actual performance. In addition, peer estimates revealed a consistent egocentric bias. High performers attributed higher accuracy to other people than did low performers. We conclude that metacognitive insight into face perception abilities is limited and subject to systematic biases. Our findings urge caution when interpreting self-report measures of face perception ability. They also reveal a fundamental source of uncertainty in social interactions.
Article
A reliable self-report measure to assess the broad spectrum of face recognition ability (FRA) from developmental prosopagnosia (DP) to super-recognition would make a valuable contribution to initial screening of large populations. We examined the performance of 96 naive participants and seven super-recognisers (SRs) using a range of face and object processing tasks and a newly developed 20-item questionnaire, the Stirling Face Recognition Scale (SFRS). Overall, our findings suggest that young adults have only moderate insight into their FRA, but those who have been previously informed of their (exceptional) performance, the SRs, estimate their FRA accurately. Principal component analysis of SFRS yielded two components. One loads on questions about low ability and correlates with perceptual tasks, and one loads on questions about high FRA and correlates with memory for faces. We recommend that self-report measures of FRA should be used in addition to behavioural testing, to allow for cross-study comparisons, until new, more reliable instruments of self-report are developed. However, self-report measures should not be solely relied upon to identify highly skilled individuals. Implications of these results for theory and applied practice are discussed.