On the Other Side of the Fence: Effects of Social Categorization and Spatial Grouping on Memory and Attention for Own-Race and Other-Race Faces

The term “own-race bias” refers to the phenomenon that humans are typically better at recognizing faces from their own than a different race. The perceptual expertise account assumes that our face perception system has adapted to the faces we are typically exposed to, equipping it poorly for the processing of other-race faces. Sociocognitive theories assume that other-race faces are initially categorized as out-group, decreasing motivation to individuate them. Supporting sociocognitive accounts, a recent study has reported improved recognition for other-race faces when these were categorized as belonging to the participants'' in-group on a second social dimension, i.e., their university affiliation. Faces were studied in groups, containing both own-race and other-race faces, half of each labeled as in-group and out-group, respectively. When study faces were spatially grouped by race, participants showed a clear own-race bias. When faces were grouped by university affiliation, recognition of other-race faces from the social in-group was indistinguishable from own-race face recognition. The present study aimed at extending this singular finding to other races of faces and participants. Forty Asian and 40 European Australian participants studied Asian and European faces for a recognition test. Faces were presented in groups, containing an equal number of own-university and other-university Asian and European faces. Between participants, faces were grouped either according to race or university affiliation. Eye tracking was used to study the distribution of spatial attention to individual faces in the display. The race of the study faces significantly affected participants'' memory, with better recognition of own-race than other-race faces. However, memory was unaffected by the university affiliation of the faces and by the criterion for their spatial grouping on the display. Eye tracking revealed strong looking biases towards both own-race and own-university faces. Results are discussed in light of the theoretical accounts of the own-race bias.     

Caucasian infants scan own- and other-race faces differently

Young infants are known to prefer own-race faces to other race faces and recognize own-race faces better than other-race faces. However, it is entirely unclear as to whether infants also attend to different parts of own- and other-race faces differently, which may provide an important clue as to how and why the own-race face recognition advantage emerges so early. The present study used eye tracking methodology to investigate whether 6- to 10-month-old Caucasian infants (N = 37) have differential scanning patterns for dynamically displayed own- and other-race faces. We found that even though infants spent a similar amount of time looking at own- and other-race faces, with increased age, infants increasingly looked longer at the eyes of own-race faces and less at the mouths of own-race faces. These findings suggest experience-based tuning of the infant''s face processing system to optimally process own-race faces that are different in physiognomy from other-race faces. In addition, the present results, taken together with recent own- and other-race eye tracking findings with infants and adults, provide strong support for an enculturation hypothesis that East Asians and Westerners may be socialized to scan faces differently due to each culture''s conventions regarding mutual gaze during interpersonal communication.     

Perceptual Other-Race Training Reduces Implicit Racial Bias

Background

Implicit racial bias denotes socio-cognitive attitudes towards other-race groups that are exempt from conscious awareness. In parallel, other-race faces are more difficult to differentiate relative to own-race faces – the “Other-Race Effect.” To examine the relationship between these two biases, we trained Caucasian subjects to better individuate other-race faces and measured implicit racial bias for those faces both before and after training.

Methodology/Principal Findings

Two groups of Caucasian subjects were exposed equally to the same African American faces in a training protocol run over 5 sessions. In the individuation condition, subjects learned to discriminate between African American faces. In the categorization condition, subjects learned to categorize faces as African American or not. For both conditions, both pre- and post-training we measured the Other-Race Effect using old-new recognition and implicit racial biases using a novel implicit social measure – the “Affective Lexical Priming Score” (ALPS). Subjects in the individuation condition, but not in the categorization condition, showed improved discrimination of African American faces with training. Concomitantly, subjects in the individuation condition, but not the categorization condition, showed a reduction in their ALPS. Critically, for the individuation condition only, the degree to which an individual subject''s ALPS decreased was significantly correlated with the degree of improvement that subject showed in their ability to differentiate African American faces.

Conclusions/Significance

Our results establish a causal link between the Other-Race Effect and implicit racial bias. We demonstrate that training that ameliorates the perceptual Other-Race Effect also reduces socio-cognitive implicit racial bias. These findings suggest that implicit racial biases are multifaceted, and include malleable perceptual skills that can be modified with relatively little training.     

Attentional Prioritization of Infant Faces Is Limited to Own-Race Infants

Background

Recent evidence indicates that infant faces capture attention automatically, presumably to elicit caregiving behavior from adults and leading to greater probability of progeny survival. Elsewhere, evidence demonstrates that people show deficiencies in the processing of other-race relative to own-race faces. We ask whether this other-race effect impacts on attentional attraction to infant faces. Using a dot-probe task to reveal the spatial allocation of attention, we investigate whether other-race infants capture attention.

Principal Findings

South Asian and White participants (young adults aged 18–23 years) responded to a probe shape appearing in a location previously occupied by either an infant face or an adult face; across trials, the race (South Asian/White) of the faces was manipulated. Results indicated that participants were faster to respond to probes that appeared in the same location as infant faces than adult faces, but only on own-race trials.

Conclusions/Significance

Own-race infant faces attract attention, but other-race infant faces do not. Sensitivity to face-specific care-seeking cues in other-race kindenschema may be constrained by interracial contact and experience.     

How Well Do Computer-Generated Faces Tap Face Expertise?

The use of computer-generated (CG) stimuli in face processing research is proliferating due to the ease with which faces can be generated, standardised and manipulated. However there has been surprisingly little research into whether CG faces are processed in the same way as photographs of real faces. The present study assessed how well CG faces tap face identity expertise by investigating whether two indicators of face expertise are reduced for CG faces when compared to face photographs. These indicators were accuracy for identification of own-race faces and the other-race effect (ORE)–the well-established finding that own-race faces are recognised more accurately than other-race faces. In Experiment 1 Caucasian and Asian participants completed a recognition memory task for own- and other-race real and CG faces. Overall accuracy for own-race faces was dramatically reduced for CG compared to real faces and the ORE was significantly and substantially attenuated for CG faces. Experiment 2 investigated perceptual discrimination for own- and other-race real and CG faces with Caucasian and Asian participants. Here again, accuracy for own-race faces was significantly reduced for CG compared to real faces. However the ORE was not affected by format. Together these results signal that CG faces of the type tested here do not fully tap face expertise. Technological advancement may, in the future, produce CG faces that are equivalent to real photographs. Until then caution is advised when interpreting results obtained using CG faces.     

Sequential Effects in Judgements of Attractiveness: The Influences of Face Race and Sex

In perceptual decision-making, a person’s response on a given trial is influenced by their response on the immediately preceding trial. This sequential effect was initially demonstrated in psychophysical tasks, but has now been found in more complex, real-world judgements. The similarity of the current and previous stimuli determines the nature of the effect, with more similar items producing assimilation in judgements, while less similarity can cause a contrast effect. Previous research found assimilation in ratings of facial attractiveness, and here, we investigated whether this effect is influenced by the social categories of the faces presented. Over three experiments, participants rated the attractiveness of own- (White) and other-race (Chinese) faces of both sexes that appeared successively. Through blocking trials by race (Experiment 1), sex (Experiment 2), or both dimensions (Experiment 3), we could examine how sequential judgements were altered by the salience of different social categories in face sequences. For sequences that varied in sex alone, own-race faces showed significantly less opposite-sex assimilation (male and female faces perceived as dissimilar), while other-race faces showed equal assimilation for opposite- and same-sex sequences (male and female faces were not differentiated). For sequences that varied in race alone, categorisation by race resulted in no opposite-race assimilation for either sex of face (White and Chinese faces perceived as dissimilar). For sequences that varied in both race and sex, same-category assimilation was significantly greater than opposite-category. Our results suggest that the race of a face represents a superordinate category relative to sex. These findings demonstrate the importance of social categories when considering sequential judgements of faces, and also highlight a novel approach for investigating how multiple social dimensions interact during decision-making.     

The Influence of Shyness on the Scanning of Own- and Other-Race Faces in Adults

The current study explored the relationship between shyness and face scanning patterns for own- and other-race faces in adults. Participants completed a shyness inventory and a face recognition task in which their eye movements were recorded by a Tobii 1750 eye tracker. We found that: (1) Participants’ shyness scores were negatively correlated with the fixation proportion on the eyes, regardless of the race of face they viewed. The shyer the participants were, the less time they spent fixating on the eye region; (2) High shyness participants tended to fixate significantly more than low shyness participants on the regions just below the eyes as if to avoid direct eye contact; (3) When participants were recognizing own-race faces, their shyness scores were positively correlated with the normalized criterion. The shyer they were, the more apt they were to judge the faces as novel, regardless of whether they were target or foil faces. The present results support an avoidance hypothesis of shyness, suggesting that shy individuals tend to avoid directly fixating on others’ eyes, regardless of face race.     

研究: 异族面孔分类中情绪的调节作用

人们对异族面孔的分类绩效比本族面孔更好,这称作异族分类优势.但面孔异族分类优势是否也和面孔异族效应一样受到情绪的调节作用,尚未可知.本文通过两个实验考察情绪对于面孔种族分类任务的影响.实验一要求40名大学生被试对不同表情面孔(中性,负性,正性)完成种族判断任务,结果发现,正性和负性情绪使年轻人对异族面孔分类变得更慢,削弱了面孔异族分类优势,而且正性情绪对异族分类效应影响更大.实验二以不同情绪加工特点的老年人为被试,发现老年人对负性异族面孔的分类绩效明显更差,负性情绪对异族分类优势的影响更大.这些结果为异族分类优势理论中,异族面孔分类和个体身份加工过程相互制约的潜在假设提供了进一步的支持.     

Own-Race Faces Capture Attention Faster than Other-Race Faces: Evidence from Response Time and the N2pc

Studies have shown that people are better at recognizing human faces from their own-race than from other-races, an effect often termed the Own-Race Advantage. The current study investigates whether there is an Own-Race Advantage in attention and its neural correlates. Participants were asked to search for a human face among animal faces. Experiment 1 showed a classic Own-Race Advantage in response time both for Chinese and Black South African participants. Using event-related potentials (ERPs), Experiment 2 showed a similar Own-Race Advantage in response time for both upright faces and inverted faces. Moreover, the latency of N2pc for own-race faces was earlier than that for other-race faces. These results suggested that own-race faces capture attention more efficiently than other-race faces.     

Adults scan own- and other-race faces differently

It is well established that individuals show an other-race effect (ORE) in face recognition: they recognize own-race faces better than other-race faces. The present study tested the hypothesis that individuals would also scan own- and other-race faces differently. We asked Chinese participants to remember Chinese and Caucasian faces and we tested their memory of the faces over five testing blocks. The participants' eye movements were recorded with the use of an eye tracker. The data were analyzed with an Area of Interest approach using the key AOIs of a face (eyes, nose, and mouth). Also, we used the iMap toolbox to analyze the raw data of participants' fixation on each pixel of the entire face. Results from both types of analyses strongly supported the hypothesis. When viewing target Chinese or Caucasian faces, Chinese participants spent a significantly greater proportion of fixation time on the eyes of other-race Caucasian faces than the eyes of own-race Chinese faces. In contrast, they spent a significantly greater proportion of fixation time on the nose and mouth of Chinese faces than the nose and mouth of Caucasian faces. This pattern of differential fixation, for own- and other-race eyes and nose in particular, was consistent even as participants became increasingly familiar with the target faces of both races. The results could not be explained by the perceptual salience of the Chinese nose or Caucasian eyes because these features were not differentially salient across the races. Our results are discussed in terms of the facial morphological differences between Chinese and Caucasian faces and the enculturation of mutual gaze norms in East Asian cultures.     

Perceptual training prevents the emergence of the other race effect during infancy

Experience plays a crucial role in the development of the face processing system. At 6 months of age infants can discriminate individual faces from their own and other races. By 9 months of age this ability to process other-race faces is typically lost, due to minimal experience with other-race faces, and vast exposure to own-race faces, for which infants come to manifest expertise [1]. This is known as the Other Race Effect. In the current study, we demonstrate that exposing Caucasian infants to Chinese faces through perceptual training via picture books for a total of one hour between 6 and 9 months allows Caucasian infants to maintain the ability to discriminate Chinese faces at 9 months of age. The development of the processing of face race can be modified by training, highlighting the importance of early experience in shaping the face representation.     

How is the individuality of a face recognized?

A familiar face is instantly recognized in a crowd. This cannot be achieved through a feature by feature comparison of the observed face with either an average face (norm-based model of face recognition) or with a set of similarly constructed faces stored in memory (exemplar-based model of face recognition). A modified norm-based model is thus proposed. Instead of memorizing an average face, the normal variations for each facial feature are used to construct a multidimensional volume of face-space devoid of unusual features, here defined as features whose metrics lie below the 5th or above the 95th percentiles for that feature. A face consisting of 100 independently variable features will thus have, on average, 10 unusual features. Face identification then becomes exception-reporting. It requires only 10 such rare features to render a given face a one in 10¹³ faces (P=0.05¹⁰=9.8×10⁻¹⁴). In a world containing 6.7×10⁹ people, such a face would be unique. Faces remembered in this way can have their unusual features exaggerated or attenuated without loss of identity. This is the basis of caricatures and anti-caricatures. It also means that individuals belonging to a foreign race, possessing several features with modes beyond the “usual range” of the own-race population, will all look alike. Features that render a face unique in the own-race population are now shared by everyone in the foreign race.Average faces are more beautiful than the faces used in the averaging process. This makes evolutionary sense. Natural selection increases the frequency of fit features at the expense of maladaptive features. “Usual features” are therefore fitter than “unusual features”, and play an important role in mate selection. Such an existing fundamental sexual attribute could easily have been harnessed for the fast and efficient recognition of individuals in the community.     

Detection of Propensity for Aggression based on Facial Structure Irrespective of Face Race

The human face provides a wealth of information pertaining to the internal state and life-stage history of an individual. Facial width-to-height ratio is a size-independent sexually dimorphic trait, and estimates of aggression made by untrained adults judging own-race faces were positively associated with both facial width-to-height ratio and actual aggressive behavior. Given the significant adaptive value of accurately detecting aggressiveness based on facial appearance, we hypothesized that aggression estimates made by adults and 8-year-olds would be highly correlated with male facial width-to-height ratio even for a face category with which they had minimal experience-other-race faces. For each of the four race and age groups, estimates of aggression were positively correlated with facial width-to-height ratio irrespective of rating own-or other-race faces. Overall, the correlations between facial width-to-height ratio and ratings of aggression were stronger for adults than for children. Sensitivity to facial width-to-height ratio appears to be part of an evolved mechanism designed to detect threats in the external environment. This mechanism is likely broadly tuned and functions independently of experience.     

Non-Threatening Other-Race Faces Capture Visual Attention: Evidence from a Dot-Probe Task

Visual attentional biases towards other-race faces have been attributed to the perceived threat value of such faces. It is possible, however, that they reflect the relative visual novelty of other-race faces. Here we demonstrate an attentional bias to other-race faces in the absence of perceived threat. White participants rated female East Asian faces as no more threatening than female own-race faces. Nevertheless, using a new dot-probe paradigm that can distinguish attentional capture and hold effects, we found that these other-race faces selectively captured visual attention. Importantly, this demonstration challenges previous interpretations of attentional biases to other-race faces as threat responses. Future studies will need to determine whether perceived threat increases attentional biases to other-race faces, beyond the levels seen here.     

Racial Bias in Neural Empathic Responses to Pain

Recent studies have shown that perceiving the pain of others activates brain regions in the observer associated with both somatosensory and affective-motivational aspects of pain, principally involving regions of the anterior cingulate and anterior insula cortex. The degree of these empathic neural responses is modulated by racial bias, such that stronger neural activation is elicited by observing pain in people of the same racial group compared with people of another racial group. The aim of the present study was to examine whether a more general social group category, other than race, could similarly modulate neural empathic responses and perhaps account for the apparent racial bias reported in previous studies. Using a minimal group paradigm, we assigned participants to one of two mixed-race teams. We use the term race to refer to the Chinese or Caucasian appearance of faces and whether the ethnic group represented was the same or different from the appearance of the participant'' own face. Using fMRI, we measured neural empathic responses as participants observed members of their own group or other group, and members of their own race or other race, receiving either painful or non-painful touch. Participants showed clear group biases, with no significant effect of race, on behavioral measures of implicit (affective priming) and explicit group identification. Neural responses to observed pain in the anterior cingulate cortex, insula cortex, and somatosensory areas showed significantly greater activation when observing pain in own-race compared with other-race individuals, with no significant effect of minimal groups. These results suggest that racial bias in neural empathic responses is not influenced by minimal forms of group categorization, despite the clear association participants showed with in-group more than out-group members. We suggest that race may be an automatic and unconscious mechanism that drives the initial neural responses to observed pain in others.     

Human Collective Intelligence under Dual Exploration-Exploitation Dilemmas

The exploration-exploitation dilemma is a recurrent adaptive problem for humans as well as non-human animals. Given a fixed time/energy budget, every individual faces a fundamental trade-off between exploring for better resources and exploiting known resources to optimize overall performance under uncertainty. Colonies of eusocial insects are known to solve this dilemma successfully via evolved coordination mechanisms that function at the collective level. For humans and other non-eusocial species, however, this dilemma operates within individuals as well as between individuals, because group members may be motivated to take excessive advantage of others'' exploratory findings through social learning. Thus, even though social learning can reduce collective exploration costs, the emergence of disproportionate “information scroungers” may severely undermine its potential benefits. We investigated experimentally whether social learning opportunities might improve the performance of human participants working on a “multi-armed bandit” problem in groups, where they could learn about each other''s past choice behaviors. Results showed that, even though information scroungers emerged frequently in groups, social learning opportunities reduced total group exploration time while increasing harvesting from better options, and consequentially improved collective performance. Surprisingly, enriching social information by allowing participants to observe others'' evaluations of chosen options (e.g., Amazon''s 5-star rating system) in addition to choice-frequency information had a detrimental impact on performance compared to the simpler situation with only the choice-frequency information. These results indicate that humans groups can handle the fundamental “dual exploration-exploitation dilemmas” successfully, and that social learning about simple choice-frequencies can help produce collective intelligence.     

Beauty Matters: Social Preferences in a Three-Person Ultimatum Game

Preference for beauty is human nature, as previous behavior studies have supported the notion of “beauty premium” in which attractive people were more easily to get promoted and receive higher salaries. In the present study, 29 males were recruited to participate in a three-person ultimatum game (UG) including a proposer, a responder and a powerless third player. Each subject, playing as the responder, had to decide whether to accept an offer from the allocator both for himself and a female third person. We aimed to elucidate how the facial attractiveness of the female subject affected the male subjects’ fairness and decision-making in social exchanges. Frontal feedback-related negativity (FRN) in response to four offers in an attractive-face condition revealed no significant differences between offers; however, when the companion was an unattractive female, an “unfair/fair” offer, which assigned a lower share to the responder and a fair share to the third player, elicited the largest FRN. Furthermore, when the third player was offered the smallest amount (“fair/unfair” offer), a larger FRN was generated in an attractive-face condition than unattractive-face condition. In the “unfair/fair” offer condition in which subjects received a smaller allocation than the third person, the beauty of their female counterparts attenuated subjects’ aversion to inequality, resulting in a less negative FRN in the frontal region and an increased acceptance ratio. However, the influence of the third player’s facial attractiveness only affected the early evaluation stage: late P300 was found to be immune to the “beauty premium”. Under the two face conditions, P300 was smallest following an “unfair/fair” offer, whereas the amplitudes in the other three offer conditions exhibited no significant differences. In addition, the differentiated neural features of processing facial attractiveness were also determined and indexed by four event-related potentials (ERP) components: N170, frontal N1, N2 and late positive potentials (LPPs).     

Faces in places: humans and machines make similar face detection errors

The human visual system seems to be particularly efficient at detecting faces. This efficiency sometimes comes at the cost of wrongfully seeing faces in arbitrary patterns, including famous examples such as a rock configuration on Mars or a toast''s roast patterns. In machine vision, face detection has made considerable progress and has become a standard feature of many digital cameras. The arguably most wide-spread algorithm for such applications (“Viola-Jones” algorithm) achieves high detection rates at high computational efficiency. To what extent do the patterns that the algorithm mistakenly classifies as faces also fool humans? We selected three kinds of stimuli from real-life, first-person perspective movies based on the algorithm''s output: correct detections (“real faces”), false positives (“illusory faces”) and correctly rejected locations (“non faces”). Observers were shown pairs of these for 20 ms and had to direct their gaze to the location of the face. We found that illusory faces were mistaken for faces more frequently than non faces. In addition, rotation of the real face yielded more errors, while rotation of the illusory face yielded fewer errors. Using colored stimuli increases overall performance, but does not change the pattern of results. When replacing the eye movement by a manual response, however, the preference for illusory faces over non faces disappeared. Taken together, our data show that humans make similar face-detection errors as the Viola-Jones algorithm, when directing their gaze to briefly presented stimuli. In particular, the relative spatial arrangement of oriented filters seems of relevance. This suggests that efficient face detection in humans is likely to be pre-attentive and based on rather simple features as those encoded in the early visual system.     

It’s All in the Eyes: Subcortical and Cortical Activation during Grotesqueness Perception in Autism

Atypical face processing plays a key role in social interaction difficulties encountered by individuals with autism. In the current fMRI study, the Thatcher illusion was used to investigate several aspects of face processing in 20 young adults with high-functioning autism spectrum disorder (ASD) and 20 matched neurotypical controls. “Thatcherized” stimuli were modified at either the eyes or the mouth and participants discriminated between pairs of faces while cued to attend to either of these features in upright and inverted orientation. Behavioral data confirmed sensitivity to the illusion and intact configural processing in ASD. Directing attention towards the eyes vs. the mouth in upright faces in ASD led to (1) improved discrimination accuracy; (2) increased activation in areas involved in social and emotional processing; (3) increased activation in subcortical face-processing areas. Our findings show that when explicitly cued to attend to the eyes, activation of cortical areas involved in face processing, including its social and emotional aspects, can be enhanced in autism. This suggests that impairments in face processing in autism may be caused by a deficit in social attention, and that giving specific cues to attend to the eye-region when performing behavioral therapies aimed at improving social skills may result in a better outcome.     

Converging Evidence of Ubiquitous Male Bias in Human Sex Perception

Visually judging the sex of another can be achieved easily in most social encounters. When the signals that inform such judgements are weak (e.g. outdoors at night), observers tend to expect the presence of males–an expectation that may facilitate survival-critical decisions under uncertainty. The present aim was to examine whether this male bias depends on expertise. To that end, Caucasian and Asian observers targeted female and male hand images that were either the same or different to the observers’ race (i.e. long term experience was varied) while concurrently, the proportion of targets changed across presentation blocks (i.e. short term experience change). It was thus found that: (i) observers of own-race stimuli were more likely to report the presence of males and absence of females, however (ii) observers of other-race stimuli–while still tending to accept stimuli as male–were not prone to rejecting female cues. Finally, (iii) male-biased measures did not track the relative frequency of targets or lures, disputing the notion that male bias derives from prior expectation about the number of male exemplars in a set. Findings are discussed in concert with the pan-stimulus model of human sex perception.