TABLE 2 - uploaded by Christian A. Meissner
Content may be subject to copyright.
Source publication
Psychological research has consistently demonstrated that individuals are better at discriminating faces of their own race when compared with faces of another, less familiar race. Given the racial/ethnic diversity of individuals screened by security personnel at transportation and border checkpoints, it is important to understand whether the cross-...
Context in source publication
Context 1
... and the perceived age difference between target images (recent age vs. distant age) on identification performance. Table 2 summarizes participants' performance across each measure. Since only ''matched'' (same person) face pairings received the perceived age difference manipulation, the findings for this effect in discrimination and response criterion are largely driven by ''hit'' rates. ...
Similar publications
Along with the increasing availability of data in many sectors has come the rise of data-driven models to inform decision-making and policy. In the health care sector, these models have the potential to benefit both patients and health care providers but can also entrench or exacerbate health inequities. Existing "algorithmic fairness" methods for...
Do people tend to be overconfident? Many think so. They’ve run studies on whether people are calibrated: whether their confidence in their opinions matches the proportion of those opinions that are true. Under certain conditions, people are systematically ‘over-calibrated’—for example, of the opinions they’re 80% confident in, only 60% are true. Fr...
In this article, we examine the state-of-the-art and current applications of artificial intelligence (AI), specifically for human resources (HR). We study whether, due to the experimental state of the algorithms used and the nature of training and test samples, a further control and auditing in the research community is necessary to guarantee fair...
It has been demonstrated that low-performing students are often inaccurate (overconfident) when predicting
their exam scores. However, these low performers are also less confident than high performers in their
predictions, indicating they may be aware of their inaccuracy. The current study had two main objectives: 1) to
replicate the findings of pr...
Citations
... The CRD is hence a statistically robust phenomenon (Lee and Penrod 2022;Malpass and Kravitz 1969;Meissner and Brigham 2001) that generalizes across faces and/or perceivers perceived as belonging to different racial groups, including individuals widely referred to as Asians, Blacks, Latinx and Whites (Chiroro et al. 2008;Simon et al. 2022;Wan et al. 2015). It has important societal implications, from eyewitness misidentifications (Brigham et al. 2007), to airport security (e.g., Meissner, Susa, and Ross 2013), to difficulties interacting with individuals perceived as cross-race (McKone et al. 2021). ...
The other-ethnicity recognition deficit (OERD) involves poorer recognition of faces perceived as other-ethnicity compared to faces perceived as same-ethnicity. In this literature, research has examined social and perceptual encoding in the OERD separately. Recent research comparing these strategies shows that encoding faces based on social information (personality traits) enhances face recognition relative to encoding faces based on perceptual information (facial features), with a similar effect for both same-ethnicity and other-ethnicity faces. Expanding on this research, we conducted three experiments (one pre-registered), manipulating the perception of face ethnicity within participants and using different social and perceptual encoding questions to examine their impacts on the OERD. Results showed that social encoding equally facilitated the recognition of faces perceived as same-ethnicity and other-ethnicity, replicating previous research (Experiment 1). However, perceptual encoding divergently impacted face recognition, improving other-ethnicity but disrupting same-ethnicity (Experiment 2) or having comparable effects to standard encoding (Experiment 3). We discuss these findings and their implications for existing OERD theories.
Keywords: face recognition, other-ethnicity recognition deficit, social encoding, perceptual encoding, face ethnicity or race
... The results often found evidence for the ORB, but especially so for White participants trying to recognise Black faces (Cross et al. 1971;Malpass and Kravitz 1969). Since the twenty-first century research exploring the ORB has employed a variety of different populations around the world and demonstrated through a number of different paradigms, such as face recognition (Hayward et al. 2017;Meissner et al. 2005;Wan et al. 2015), face matching tasks (Havard 2021;Kokje et al. 2018;Meissner et al. 2013), eyewitness paradigms (Dodson & Dobolyi 2016;Havard et al. 2017;Jackiw et al. 2008;Marcon et al. 2008), and metaanalyses (Meissner and Brigham 2001;Singh et al. 2021). Most studies have found that people are more likely to correctly identify a previously seen face if it belongs to the same race as them and more likely to make a false positive response (falsely recognise a face) if it belongs to someone who is from a different race. ...
... The results often found evidence for the ORB, but especially so for White participants trying to recognise Black faces (Cross et al. 1971;Malpass and Kravitz 1969). Since the twenty-first century research exploring the ORB has employed a variety of different populations around the world and demonstrated through a number of different paradigms, such as face recognition (Hayward et al. 2017;Meissner et al. 2005;Wan et al. 2015), face matching tasks (Havard 2021;Kokje et al. 2018;Meissner et al. 2013), eyewitness paradigms (Dodson & Dobolyi 2016;Havard et al. 2017;Jackiw et al. 2008;Marcon et al. 2008), and metaanalyses (Meissner and Brigham 2001;Singh et al. 2021). Most studies have found that people are more likely to correctly identify a previously seen face if it belongs to the same race as them and more likely to make a false positive response (falsely recognise a face) if it belongs to someone who is from a different race. ...
In police photo lineups, there can sometimes be small variations in shades and hues of the background images due to the faces being filmed under different lighting and cameras. Own race bias refers to a situation where people are better at remembering the faces of those who are the same race as them and find it more difficult to recognise faces from a different race. In this paper, we investigated the influence of small colour variations in backgrounds for the recognition of Black and White faces. Across 3 experiments, we found when small changes were introduced into the backgrounds of the images this increased false identifications for previously unseen Black faces, but not White faces. This finding suggests that the police need to ensure that the backgrounds of the photo lineups they use are all uniform to reduce mistaken identifications of innocent suspects.
... Concerningly, it is possible that these issues might be compounded by ethnicity, both of the human operator and the faces being examined. With reports of racial biases in some AFRS (Grother et al., 2019;Phillips, Jiang, et al., 2011), and other race effects (Meissner & Brigham, 2001) documented in the unfamiliar face matching performance of humans (Megreya et al., 2011;Meissner et al., 2013), further research is needed to examine how human-AFRS teams perform when matching faces of various ethnicities. ...
Automated Facial Recognition Systems (AFRS) are used by governments, law enforcement agencies, and private businesses to verify the identity of individuals. Although previous research has compared the performance of AFRS and humans on tasks of one-to-one face matching, little is known about how effectively human operators can use these AFRS as decision-aids. Our aim was to investigate how the prior decision from an AFRS affects human performance on a face matching task, and to establish whether human oversight of AFRS decisions can lead to collaborative performance gains for the human-algorithm team. The identification decisions from our simulated AFRS were informed by the performance of a real, state-of-the-art, Deep Convolutional Neural Network (DCNN) AFRS on the same task. Across five pre-registered experiments, human operators used the decisions from highly accurate AFRS (> 90%) to improve their own face matching performance compared with baseline (sensitivity gain: Cohen's d = 0.71-1.28; overall accuracy gain: d = 0.73-1.46). Yet, despite this improvement, AFRS-aided human performance consistently failed to reach the level that the AFRS achieved alone. Even when the AFRS erred only on the face pairs with the highest human accuracy (> 89%), participants often failed to correct the system's errors, while also overruling many correct decisions, raising questions about the conditions under which human oversight might enhance AFRS operation. Overall, these data demonstrate that the human operator is a limiting factor in this simple model of human-AFRS teaming. These findings have implications for the "human-in-the-loop" approach to AFRS oversight in forensic face matching scenarios. (PsycInfo Database Record (c) 2022 APA, all rights reserved).
... Paul McCartney, when he was in the Beatles, looks very different to his present appearance. Whereas matching identity in unfamiliar faces is especially challenging when the age gap between two images increases (Davis & Valentine, 2009;Megreya et al., 2013;Meissner et al., 2013), age-related changes to appearance do not impair recognition of familiar faces. For example, in the Yearbook Task, participants were remarkably accurate in their ability to match photographs of their former classmates from their high school yearbook to images taken 24 to 26 years later (Bruck et al., 1991). ...
Matching identity in images of unfamiliar faces is error prone, but we can easily recognize highly variable images of familiar faces – even images taken decades apart. Recent theoretical development based on computational modelling can account for how we recognize extremely variable instances of the same identity. We provide complementary behavioural data by examining older adults’ representation of older celebrities who were also famous when young. In Experiment 1, participants completed a long‐lag repetition priming task in which primes and test stimuli were the same age or different ages. In Experiment 2, participants completed an identity after effects task in which the adapting stimulus was an older or young photograph of one celebrity and the test stimulus was a morph between the adapting identity and a different celebrity; the adapting stimulus was the same age as the test stimulus on some trials (e.g., both old) or a different age (e.g., adapter young, test stimulus old). The magnitude of priming and identity after effects were not influenced by whether the prime and adapting stimulus were the same age or different age as the test face. Collectively, our findings suggest that humans have one common mental representation for a familiar face (e.g., Paul McCartney) that incorporates visual changes across decades, rather than multiple age‐specific representations. These findings make novel predictions for state‐of‐the‐art algorithms (e.g., Deep Convolutional Neural Networks).
... The ORB has been replicated using a variety of paradigms including face matching tasks, where participants have to decide if two images are the same person or different individuals. Findings from these simultaneous matching tasks report that people are more accurate at matching own-race faces, when compared with other-race faces (Kokje et al., 2018;Meissner et al., 2013). Kokje et al. (2018) presented Caucasian and Egyptian participants with own and other-race faces for a matching task and found not only an ORB, but that matching was more accurate for matched pairs, when compared with mismatched pairs. ...
... Kokje et al. (2018) presented Caucasian and Egyptian participants with own and other-race faces for a matching task and found not only an ORB, but that matching was more accurate for matched pairs, when compared with mismatched pairs. Meissner et al. (2013) also replicated the ORB when they presented Mexican American (own race) and African American (other race) target faces along with passport photographs to Mexican American participants and found greater accuracy for own-race pairings. ...
Research has shown that we are better at discriminating between faces that are our own race, and much less accurate with faces of another race. When the external features of faces were removed, this reduced the accuracy for recognizing other-races faces, more than own-race faces, suggesting that the external features (hair, face shape) are especially important for the recognition of other-race faces. The aim of the current study was to determine whether external features were more useful in matching other-race faces, and whether this was the case for Western and Eastern viewers. The current study employed a face matching task with Caucasian (U.K.) and Asian (Chinese) participants and found that responses were more accurate for own-race faces, and for whole faces when compared with faces where the internal or external features had been removed. Removing the external features of other-race faces increased the own-race bias for Chinese and U.K. participants, demonstrating the importance of viewing whole faces, including the external features when matching other-race faces.
... However, similarly poor levels of performance are found in studies of viewers matching a photo to a live person in front of them (e.g., Kemp et al., 1997;Megreya & Burton, 2008;White et al., 2014). Furthermore, matching tasks involving images embedded in ID documents generally give rise to similar overall levels of accuracy as tasks using isolated face images (Bindemann & Sandford, 2011;Kramer et al., 2019;Meissner et al., 2013). ...
Matching unfamiliar faces is a well-studied task, apparently capturing important everyday decisions such as ID checks. In typical lab studies participants make same/different judgements to pairs of faces, presented in isolation and without context. However, it has recently become clear that matching faces embedded in documents (e.g. passports and driving licences) induces a bias, resulting in elevated levels of ‘same person’ responses. While practically important, it remains unclear whether this bias arises due to expectations induced by the ID cards or interference between textual information and faces. Here we observe the same bias when faces are embedded in blank (i.e. non-authoritative) cards carrying basic personal information, but not when the same information is presented alongside a face without the card (Experiments 1 & 2). Cards bearing unreadable text (blurred or in an unfamiliar alphabet) do not induce the bias but those bearing arbitrary (non-biographical) words do (Experiments 3 and 4). The results suggest a complex basis for the effect, relying on multiple factors which happen to converge in photo-ID.
... In a previous study [32], it was shown that forensic facial compar- Comparing the outcomes of the present study to published literature is difficult as no other studies followed a feature-based systematic morphological analysis, and only a few of the studies investigating disguises in face matching reported error rates (e.g., [15,50]). Most studies tested the performance of multiple observers, usually members of the public and not trained experts, without using a systematic approach (e.g., [51]). Nonetheless, in their experiment involving members of the public attempting to identify the presence of a target with a brimmed cap from CCTV footage, Henderson et al. [15] found markedly high false-positive rates, ranging between 40% ...
... General agreement exists in the literature that there is a decrease in ability to recognize faces which are partially disguised by hats and eyeglasses [14,15,17,[21][22][23]53]. However, many of these studies address disguises with a focus on specific settings of passport security screening (e.g., [51,54]) or impersonation (e.g., [53]), often with little consistency in the actual disguise accessories used (e.g., [53] The superior performance of morphological analysis using a systematic approach in the sunglasses cohort over the brimmed cap cohort is in agreement with other studies. Lee et al. [14] identified the lowest accuracy in their sample of headgear with a brim (57.1%) as well as Henderson et al. [15] who identified even lower accuracies (between 40 and 43%) in their brimmed hat face-matching task. ...
... Expectedly, the decrease in face recognition ability has been found to be further exacerbated when both caps and glasses are included in face recognition tasks [51]. Meissner et al. [51] identified a notable decrease in discriminatory accuracy (a computed accuracy that accounts for false positives, but not false negatives) ranging from 57% to 73% in disguised comparisons [51]. ...
Disguises are commonly used to mask a person's facial appearance in areas under closed-circuit television (CCTV) surveillance. While many studies attempted to understand the effects of disguises, such as hats and glasses, on facial recognition, limited studies have looked at disguises in forensic facial comparison. The aim of this study was to compare the outcomes of forensic facial comparison by morphological analysis (MA) in a CCTV sample with sunglasses and brimmed caps. The sample was obtained from the Wits Face Database and organized into 81 face pools of one target facial image wearing a disguise (cap or sunglasses) and 10 potential matching images. MA was conducted across face pools, and confusion matrices were used to assess the outcomes. Surprisingly, sunglasses had limited effect on MA performance both in accuracy (90.4%) and in reliability (κ = 0.798), while caps markedly decreased both accuracy (68.1%) and reliability (κ = 0.639). Error rates were associated primarily with false negatives in both samples (caps: 42.4%; sunglasses: 16.1%) despite the sample distribution favoring false-positive errors, which were very low (caps: 0.6%; sunglasses: 0%). Similarly to other studies, hats and caps were more harmful to correct identification when compared to sunglasses, which actually resulted in better accuracy than regular CCTV recordings. The effect of brimmed caps on accuracy was attributed to the overall loss of facial information caused. On training analysts, it may be helpful to instruct purposefully avoiding overreliance on easily disguised facial features, as other regions of the face also contain substantial feature information.
... The improved recognition of own-race faces compared to other-race faces is known as the cross-race effect (Hancock & Rhodes, 2008;Hayward, Crookes, & Rhodes, 2013). Reduced recognition of other-race faces is problematic in many real-world situations such as social interactions, eyewitness testimony, and person identification at airports and border checkpoints (Behrman & Davey, 2001;Meissner, Susa, & Ross, 2013). Most studies investigating the cross-race effect have used long-term recognition memory paradigms (Meissner & Brigham, 2001). ...
Previous research has shown that working memory (WM) performance for own‐race faces is better than for other‐race faces. We focused on the storage capacity and encoding rate to identify WM characteristics that facilitate own‐race face recognition. We investigated WM's temporal dynamics for own‐ and other‐race faces to separately identify the contribution of storage capacity and encoding rate on the own‐race advantage in WM. We presented Asian participants with Asian faces as own‐race faces and Black faces as other‐race faces in two experiments. Experiments 1 and 2 indicated a higher storage capacity for own‐race faces, and Experiment 2 also indicated an increased encoding rate for own‐race faces when backward masking was used. Moreover, there was no association between storage capacity and encoding rate. These findings suggest that both storage capacity and encoding rate independently contribute to the cross‐race effect in WM.
... To investigate this possibility, we examined ensemble coding of unfamiliar faces from two face categories with which adults lack experience: other-race faces and inverted faces. Adults show an own-race advantage when asked to recognize images of unfamiliar faces (e.g., Golby, Gabrieli, Chiao & Eberhardt, 2001;MacLin & Malpass, 2001;Meissner & Brigham, 2001;Wright, Boyd & Tredoux, 2003), detect changes in feature shape and spacing (Hayward, Rhodes, & Schwaninger, 2008;Mondloch et al., 2010;Rhodes et al., 2009;Zhao et al., 2014), match identity across images of wholly unfamiliar faces (Laurence, Zhou & Mondloch, 2016;Megreya, White, & Burton, 2011;Meissner, Susa & Ross, 2013;Proietti, Laurence, Matthews, Zhou & Mondloch, 2019), and learn a newly encountered face based on low variability in appearance (Zhou, Matthews, Baker & Mondloch, 2018); all of these effects have been attributed to a lack of experience Kelly et al., 2005Kelly et al., , 2007Tanaka, Heptonstall & Hagen, 2013). Likewise, adults show an upright face advantage when asked to recognize familiar faces (Scapinello & Yarmey, 1970;Yarmey, 1971), when asked to match identity of unfamiliar faces (Megreya & Bindemann, 2015;Megreya & Burton, 2006) and on tasks that assess holistic processing and sensitivity to feature spacing (Freire et al., 2000;Mondloch, Le Grand & Maurer, 2002;Thompson, 1980;Young, Hellawell, & Hay, 1987). ...
The ability to recognize identity despite within‐person variability in appearance is likely a face‐specific skill and shaped by experience. Ensemble coding – the automatic extraction of the average of a stimulus array – has been proposed as a mechanism underlying face learning (allowing one to recognize novel instances of a newly learned face). We investigated whether ensemble encoding, like face learning and recognition, is refined by experience by testing participants with upright own‐race faces and two categories of faces with which they lacked experience: other‐race faces (Experiment 1) and inverted faces (Experiment 2). Participants viewed four images of an unfamiliar identity and then were asked whether a test image of that same identity had been in the study array. Each test image was a matching exemplar (from the array), matching average (the average of the images in the array), non‐matching exemplar (a novel image of the same identity), or non‐matching average (an average of four different images of the same identity). Adults showed comparable ensemble coding for all three categories (i.e., reported that matching averages had been present more than non‐matching averages), providing evidence that this early stage of face learning is not shaped by face‐specific experience.
