The Other Race-Effect: A Brief Review

In light of the recent and on-going cases of overt violence and discriminatory behaviours towards segregated populations, I felt that sharing this piece was appropriate and necessary. This is a very wide ranging topic with many co-factors so I tried my best to cut what wasn’t necessary and keep what is. I believe education is one of the few real ways these negative interactions will stop, and it all starts with us. Understanding the neuropsychological parameters around interracial interactions, specifically between different races who are not well mixed, in my opinion, is a potent source to derive solutions from. I intended to maintain an objective perspective as best I could while reporting the research up to this point. With that said, if anything in here seems biased, incorrect, or if you have any questions, please don’t hesitate to leave a comment. I hope some of you can draw something from this research.

Thoughts and social behaviours are highly influenced by previous social interactions with other individuals (Teles et al., 2016).  Many types of social interactions include an exchange of emotional information that can be conveyed through the content of a message or the tone of voice, all of which contribute to the gestalt of social exchange (McColl & Nejat, 2014).  The expression of emotion, specifically through facial expressions, has a significant impact on the emotional content of an interaction.  However, the interpretation of an emotional message can be influenced by a number of factors, such as where that individual comes from, or what kind of learned experiences an individual has had (Gendron et al., 2018).  Culture, typically defined as a large group of people who more or less share similar types of learned experiences, has been shown to influence the individual not only in the perception of emotional expression, but in the perceived meaning of an emotion.  Various races (e.g., White, Black, etc.) demonstrate cultural norms that differ from other cultures and races such as greetings, facial expression meanings, and discrete interaction norms.  For example, research by Gendron and colleagues (2018) showed that individuals living in small-scale societies in Africa and South America demonstrate a difference in perceived meaning of facial expressions when compared to westernized populations.  Individuals from these communities were more likely to attribute certain facial expressions to an explicit behaviour such as looking or smelling, rather than an expression that represents an internal mental event such as feeling angry or sad, while United States (U.S) participants demonstrated the opposite (Gendron et al., 2018).  These findings suggest that this difference in perception and intended meaning of a facial expression can be culturally influenced and is thus reinforced as a mode of social perception.  This difference in meaning of facial expressions from one culture to the next informs researchers of the effect of subtle, individuating experiences and how they lead to lasting changes in visual and emotional processing, behaviour, and ultimately, cultural norms (Curby et al., 2019; Gendron et al., 2018; Kubota & Ito, 2014).  A psychosocial consequence of these differences in perception across race can lead to negative social outcomes such as racial discrimination that stem from perception informed implicit biases.  Through a deeper understanding of experience sensitive perceptual systems, researchers are offered a new perspective into what may drive or influence behaviours, thoughts, and how they are expressed.  This review intends to explore if there is a difference in the way individuals process emotional faces of same-race (SR) and other-race (OR) individuals in a culturally-modulated, possibly non-conscious, direct processing pathway.  Through the exploration of possible visual streams that contain adaptive, plastic processing characteristics that seem independent of the more traditional bottom-up, component-based processing stages suggests that there could be visual streams that are unique to facial emotion.  The findings obtained from past research will help identify potential direct processing streams through which facial expressions are recognized prior to the discrete local processing of individual facial elements that traditionally defined perceptual processing.  These research findings can provide important insights into our understanding of social avoidance, cultural/racial perception bias, Autism Spectrum Disorder, cortical plasticity, and even in clinical neurological conditions such as prosopagnosia.  A deeper understanding of these more direct, special case streams could invite the possibility of ameliorating the associated negative side effects that develop from cultural variation within facial emotion perception. 

Visual Processing

The human face as a stimulus is believed by many researchers to be processed holistically, particularly in neuroimaged cortical areas such as the Fusiform Face Area (FFA) and Occipital Face Area (OFA) (Richler & Gauthier, 2014; Schiltz & Rossion, 2006).  The FFA is believed to be populated with neurons that are sensitive to face identities, which respond strongly to new faces and adapt in response to repetitive viewing of the same identities.  Holistic processing is a mode of perception whereby multiple components of stimuli are processed in a parallel fashion, effectively reducing the ability to selectively attend to specific features (Hancock & Rhodes, 2008).  When an individual incurs damage in areas associated with facial processing, symptoms such as face blindness, or prosopagnosia can occur (Wan, et al, 2017).  More generally, holistic processing is believed to partially occur in the FFA and OFA and is thought to be a product of stimulus familiarity such as the arranged components of a human face, or in the case of non-faces, familiarity through exposure time or an in-depth understanding of various categories of objects (Bukach et al., 2012; Feng et al., 2011).  As familiarity increases as a product of frequent exposure, there is a reflective drop in reaction time in recognition tasks due to the reduced reliance on bottom-up, component based processing and an increase in parallel processing of familiar facial features (Hancock & Rhodes, 2008; Kubota & Ito, 2014).  Curby et al. (2019) has suggested that there are unique holistic processing differences in regards to a human face that have not been seen in high-expertise object recognition tasks, which suggests face-specific processing streams.  As holistic processing increases (i.e., an increase in parallel processing) as a result of exposure to a given stimulus (i.e., a human face), discriminability and recognition of differences increases (Curby et al., 2019).  Interestingly, a meta-analysis conducted by Sugden & Marquis (2017) suggests that the ability to discriminate between emotional expressions is a facility gained within hours after birth.  From early childhood (6-9 months), infants will display marked response differences (i.e., looking longer or shorter) to faces of varying emotional saliency.  This early display of sensitivity to various emotional expressions suggests that there may be special case recognition visual streams as early as 0-3 months old that develop rapidly.  The capacity to discriminate between facial expressions such as happy and angry improves over time as the visual system develops and is influenced by individuating experiences, not reaching maximal sensitivity until middle age (Sugden & Marquis, 2017).  The ability to discriminate between happy and angry facial expressions seems to develop equally; however, some research suggests that there is a perceptual bias given to angry faces over happy faces (Sugden & Marquis, 2017).  Using psychophysical masking techniques, Wesner et al. (2015) explored the processing of facial emotions by examining the hierarchical nature of discriminating angry or happy faces from neutral ones.  It was found that discriminability can occur subceptively (i.e., below conscious levels), with significantly shorter response times and greater efficiency rates for angry over happy drawn faces.  Furthermore, this was true not only for upright drawn facial emotions, but for inverted ones as well, contrary to earlier neuroimaged findings (Bukach et al., 2012).  However, presenting identically drawn curving features found within the faces in isolation showed no effect.  This finding supports the contextual, holistic processing characteristics of facial emotion, with enhanced response times and accuracy for context angry expressions, even when the emotional intensity, as a function of log arc degrees of mouth and eyes (curvature of mouth and eye as a proxy for emotional intensity), was so subtle in distinction that the observer is not consciously aware of an emotion.  With this in mind, it is plausible that societal or cultural processing influences can occur within an attentional “top-down” model; however, Wesner et al. (2015) explored the processing of potentially threatening stimuli at the sub-threshold level and found that based on a masking technique, the facial emotion discriminations were not attributable to cognitive avoidance.  Cognitive avoidance is the tendency to orient attention away from negative stimuli (Sagui, 2017).  Similar to what happens with object expertise, it is thought that subliminal discrimination differences with respect to the processing of threatening angry faces might be observed when participants are encountering faces that are stereotypically categorized as SR versus OR faces that are unfamiliar to a participant based on low prior social experience.  Alternatively, if cognitive avoidance is not influencing the detection of threatening facial emotions, then perhaps there are extraneous causes beyond a potential social-cognitive component. 

Other-Race Effect 

The impact of familiarity has been demonstrated when individuals perceive faces of their own racial group versus faces from other less familiar racial groups, with better recognition capabilities being shown for the same race as the viewer (Anthony et al., 1992; Bothwell et al., 1989; Malpass & Kravitz 1969).  This facial recognition difference is known as the Other Race Effect (ORE) (Bukach et al.,2012; Hart et al., 2000; Herzmann et al. 2011; Tanaka et al., 2004).  The ORE is a well known component of social neuroscience that is thought to be a product of a lack of exposure and individuating experiences with another race (Zhou, 2018).  More generally, although research demonstrates various probable causes for the ORE, it is generally taken to be a product of one’s contextual social surroundings (e.g., frequency of contact; Brigham et al., 1982) and personal history (Bukach, et al., 2012).  Through studies done on self-reported levels of contact with other races, those who had high levels of contact demonstrated a smaller ORE (i.e, increased parallel processing; Hancock & Rhodes, 2008).  The reduction in ORE is thought to be indicative of changes in the processing of OR faces in response to repeated exposure.  Evolutionarily, the ORE can be speculated to have been an adaptive mechanism keeping an individual within their respective tribe or society with an increased capacity to read subtle emotional cues given by local individuals.  Infact, research shows that the ORE is present in face-recognition tasks with infants as early as 3 months old (Sugden & Marquis, 2017).  Consequently, it can be speculated that holistic processing of OR faces is a perceptual privilege gained only after a requisite amount of experience with OR faces, while SR holistic processing is developed within months after birth.  Based on the idea that we process SR faces more holistically, the ORE may be a product of facial processing characteristics that are strongly associated with cultural, societal, and/or individuating influences (Zhou, 2018).  Infact, fMRI research has shown that holistic processing characteristics, specifically within the right FFA, differ when perceiving SR faces versus OR faces.  Noted differences show a neural composite face effect only for SR faces in the right FFA, with a release of adaptation in response to OR faces (Zhou, 2018).  A neural composite effect refers to the tendency for individuals to recognize misaligned faces (i.e., top half does not align with the bottom half of a face) better than aligned faces.  This is thought to be an indication of holistic processing because the misaligned bottom and top of a face are recognized as a new identity, which is fused and processed together within the right FFA (Zhou, 2018).  These marked neurophysiological differences when perceiving SR and OR faces suggests experience sensitive, racially defined recognition processing differences (Zhou, 2018).  Zhou and colleagues (2018) findings are bolstered by additional research that found through Event-Related Potentials (ERP), a difference in processing of White participants viewing OR (i.e., “Black”) faces which evoked larger N100s and P200s than did “White” faces, while “White” faces evoked larger N200s than did “Black” faces (He et al., 2009).  Notably, race differences for faces can be detected as early as 116 milliseconds (ms) with a reflective difference in resultant ERP activity (He et al., 2009).  Researchers theorize that the time sequences indicate a rapid capture of attention in response to OR faces, while SR faces undergo extended processing which may explain the better recognition abilities for SR faces.  

This difference in processing poses as a real world obstacle.  Wan and colleagues (2017) suggests that some groups of people could meet criteria for clinical level impairment due to their inability to perceive differences between faces of OR individuals, rendering the individual effectively face-blind for other races.  An example of an ORE-derived social obstacle is perceived rudeness or racism following a perceptual failure to recognize a member of another race which can increase amounts of perceived racism for OR individuals.  Although only 2% of the population show OR face-blindness,  this small number can still have a substantial impact on both the perceived individual (e.g., racism) and the perceiving individual (e.g., perceived memory problems; Wan et al., 2017).  Findings by Wan and colleagues (2017) suggest that individuals with very low experience with OR individuals, namely in legal settings like eye witness testimonies, pose as a real threat to misidentification.  A significant sample of literature also suggests that this reduced performance in visual processing of OR individuals can lead to or influence shifts in implicit attitudes towards those OR individuals (Bukach et al., 2012; Kubota & Ito, 2014; Wan et al, 2017).  Interestingly, the ORE can be reduced through individuating experiences with repetitive exposure to OR faces (Wan et al., 2017). 

Implicit Bias

Implicit bias is the psychosocial consequence of culturally reinforced stereotypes towards “outgroup” or stigmatized individuals that influence aspects of social neuroscience ranging from overt behaviour such as avoidance (e.g., perceived danger; Kubota & Ito, 2014) to cognitive processing (i.e., differences in facial processing; Anthony et al., 1992; Bukach et al., 2012; Kubota & Ito, 2014).  Implicit bias is inherently non-conscious, rather than explicit bias, which would be an attitude or bias that an individual is consciously aware of.  Although the neurobiological characteristics of implicit bias are being uncovered, implicit bias is still an important area of research as it is thought to play a role in social prejudicial behaviour that may be reminiscent of significant cultural and/or perceptual biases which deserve exploration (Kubota & Ito, 2014).  Prejudicial behaviour, such as discrimination, can be driven by subconscious associations reminiscent of observations of stereotypical behaviour that may also be reinforced by societal attitudes (Bukach et al., 2012).  The attribution of negative stereotypes to certain groups of people has been demonstrated to serve as an obstacle for transcultural relations and thus deserves scientific evaluation.  

Implicit Association Test

The non-conscious perceptual filter of implicit bias is typically measured by the Implicit Association Test (IAT; Greenwald, et al., 1998; Implicit Association Test, 2011).  The Implicit Association Test (Greenwald et al., 1998) is a reaction time-based test used to evaluate the strengths of subconscious implicit biases towards various categories.  Reaction times (in ms) are recorded to determine the categorization speed of the participants and are used to determine implicit bias.  Reaction time is used because latency rates are thought to be a proxy for the strength of a given association and increased processing of stimuli, with more familiar associations being more rapidly activated (Hancock & Rhodes, 2008).  There are many categories the IAT can examine such as skin-tone, religious affiliation, body weight, sexual preference etc.  To explore an individual’s subconscious attitude towards members of another race, researchers use the Skin-Tone or Race IAT, which includes photographs of White, Black, and/or Asian individuals.  

The test consists of an individual being presented with two categories assigned to specific keys on a keypad that are used to categorize a presented stimulus such as a connotated word (e.g., disgust or spectacular) or picture (e.g., human face) under one of the two categories (e.g., “good” or “bad”/“White” or “Black”).  The beginning of the test includes two categories denoted “good” and “bad” in the top corners of the screen.  The individual is presented with a positively or negatively connotated word in the centre of the screen, which they are to categorize under one of the two categories given.  Following that set of trials, categories switch (“good, “bad” switch keys and spots on the screen) so as to reduce repetition and practice effects.  The categories then switch from “Bad” and “Good” to “Black” and “White”; the stimulus is also changed from connotated words to images of faces.  The conditions repeat with the addition of coupled categories (“Good or White”, “Bad or Black”) and the individual is to categorize the presented stimulus which includes either a connotated word or a picture of a “Black” or “White” face.  

Theoretically, personally held associations that are congruent with the presented stimulus will be more rapidly and reliably chosen (Greenwald et al., 1998).  For example, an individual who holds negative attitudes towards Black individuals will be more likely to associate “Black” and “bad”, which will be demonstrated through lower reaction times for the “Black or Bad” trials.  Various adaptations of IAT measures usually reach internal consistency estimates (split-half correlations or Cronbach’s alphas) between 0.70 and 0.90 (Schnabel, Asendorpf, & Greenwald, 2008).  This is a psychometrically satisfactory estimate according to most researchers.  Test–retest reliability has been observed to show a median of 0.56 across different studies which is about 0.15 to 0.20 below the internal consistencies that are typically obtained for IAT measures (Schnabel, Asendorpf, & Greenwald, 2008).  In response to the psychometric properties of the IAT, some implications that have been drawn from results have come under questioning (Blanton & Mitchell, 2011).  Researchers have noted that although the IAT is useful in exploring associations between concepts, the public (and some researchers; Greenwald et al., 1998) has suggested that the results indicate a more meaningful representation of explicit attitudes and social behaviour than the research would suggest.  More specifically, although an individual may demonstrate an association that suggests a preference for their own race over another, the research has yet to demonstrate a significant resultant difference in explicit attitudes and behaviours towards OR individuals (Blanton & Mitchell, 2011).  

Implications of Implicit Bias

Implicit bias has been found to influence the processing of faces of other races (Hancock and Rhodes, 2008).  The difference in processing speeds of OR vs. SR facial emotional expressions are thought to be manifest of the ORE (i.e., frequency of exposure) and the resultant difference in holistic processing characteristics as mentioned above.  Hancock and Rhodes (2008) delineate relevant components of processing that are influenced by implicit bias in OR facial processing such as differences in the way facial components are holistically processed, configural coding and featural coding.  Apparently, these components of processing, particularly holistic processing, are sensitive to lasting changes due to individuating experiences that are repetitive and variant (Zhou et al., 2018).  Individuating experiences are a product of stimulus exposure frequency (Bukach et al., 2012).  Furthermore, the more frequent a particular stimulus is perceived, the better the recognition efficiency is for that particular stimulus due to an increase in holistic processing.  The idea that individuating experiences lead to increased identification efficiency is supported by the finding that familiarity with OR individuals positively correlates with performance in face recognition tasks (Bukach et al., 2012).  Thus, differences in facial recognition across races suggest reflective differences in processing characteristics, that are as the research suggests, influenced by experience.  These holistic processing characteristics are not only impacted by personal experiences, but are also reactive to contextual changes (e.g., a facial expression, contextual surroundings, internal states; Kubota & Ito, 2014).  Various emotional expressions such as sadness or happiness can influence the effect of implicit bias.  Research by Kubota and Ito (2014), sought to explore the effects of happy, angry, or neutral faces of “White” and “Black” people on a weapons/tools categorization task done by White participants.  Through the measurement of response latency and error rates, Kubota and Ito (2014) found that facial expression modulated implicit stereotyping.  Results showed that a “Black” angry face led to an increased implicit stereotypical response where the response was in accordance with the ORE: angry faces of another race elicit threat responses, while happy faces displayed a reduction in implicit stereotype (Kubota & Ito, 2014).  Furthermore, responses following neutral facial expressions demonstrate a functional reliance on context whereby threatening facial expressions can modulate the perceived threat from other neutral faces.  This finding was displayed in a study where a “Black” neutral face presented after a “Black” angry face elicited crime and danger-relevant stereotypes (Kubota & Ito, 2014).  Researchers concluded that because different associations (e.g., “Black” angry face and threat) are activated based on changes in emotional expression, positive facial expression components such as a smile may attenuate implicit bias (Kubota & Ito, 2014).  The internal state of the perceiver has also been shown to have a significant influence on implicit bias.  Lee and colleagues (2018) demonstrated that conceptualization of internal negative affect, particularly in this case as sympathy or fear, had a significant impact on implicit bias.  Evidence found that the conceptualization of negative affect toward Black Americans as sympathy, rather than fear, reduced racial bias (Lee et al., 2018).  For example, White Americans who feel fearful toward Black Americans are more likely to oppose government policies that seek to aid Black People.  Alternatively, White Americans who felt sympathy towards Black Americans were more likely to support policies like affirmative action (Lee et al., 2018).  Research demonstrates that negative affect is not always indicative of discriminatory behaviour, but rather the knowledge that a group has experienced oppression historically.  Lee et al., (2018) highlighted that negative affect can be perceived due to the knowledge that a group has gone through negative experiences which suggests that negative affect is not always indicative of animosity towards outgroup members, conscious or otherwise, but rather a failure to properly conceptualize an emotion.  Implicit bias can be highly attenuated by reducing emotions associated with fear and antisocial behaviours (anger, disgust) and by increasing prosocial emotions (sympathy, love, happiness; Lee et al., 2018).  Researchers have also successfully attenuated implicit bias through meditation.  A study done by Kang and colleagues (2014) demonstrated that a loving-kindness meditation program reduced implicit bias towards societally stigmatized populations.  Loving kindness meditation is intended to cultivate warm and friendly feelings toward the self and others.  Roughly 100 non-Black, non-homeless participants were recruited and randomly assigned to one of three conditions consisting of a six-week loving-kindness meditation practice group, a six-week loving-kindness discussion group, and a waitlist control.  Participants assigned to the six-week loving-kindness meditation group were the only participants to experience a reduction in automatic implicit bias towards Blacks and homeless people as measured by the IAT.  Kang et al., (2014) concluded that a loving-kindness meditation practice reduces automatically activated implicit bias through increased conscious control suggesting that the cognitive context of the participant influences automatic implicit bias activation.  Kang and colleagues (2014)  also noted that a simple relaxation exercise did not demonstrate this finding, suggesting that a common byproduct of meditation, stress reduction, is not responsible for the reduction in implicit bias towards Blacks, and only marginally influenced implicit bias towards homeless people.  From these studies, it is apparent that implicit bias is highly sensitive to the internal state of the perceiver. 

If implicit bias leads to objective, lasting changes in visual processing, there should also be changes observable beyond behavioural measurements.  Research conducted by He and colleagues, (2009) demonstrated differences in ERP recordings between Asian, Black, and White participants on a gender identification task.  Researchers correlated electroencephalography (EEG) recordings with IAT results to find psychophysically supported implicit bias.  ERP differences between SR and OR faces were strongly correlated with differences in implicit attitudes measured by the IAT.  Signals peaked at different points for each race; late positive component at 592 ms was greater for SR faces, which the researchers concluded is manifest of deeper, extended processing of SR faces.  This finding is somewhat at odds with findings from previous studies that suggest SR faces are processed more rapidly than OR faces.  However, the extended processing of SR faces may be unrelated to implicit associations and may be indicative of further visual processing that relate to memory or emotional connectedness (He et al., 2009). 

These findings suggest several factors can affect the relationship between implicit bias and the processing of OR facial emotions.  Research has demonstrated effects of: internal and external context (Kang et al., 2014; Lee et al., 2018), past experience with other races (Hancock & Rhodes, 2008), and attitudes towards stigmatized populations (Bukach et al., 2012).  Past research has demonstrated the effect of implicit bias on OR facial expression processing through a mediation of gestalt and component based processing of facial features (Gendron et al., 2018).  Existing research suggests that implicit bias is mediated by the amount of exposure to another race an individual has had across a lifetime.  However, there has yet to be an investigation of the exact mechanism implicit bias influences in regards to the processing of facial emotions of various intensities.  As described earlier, direct recognition and processing streams that may be unique to facial expressions with respect to implicit bias have yet to be clearly defined.  

I hope you managed to glean something out of this. Below I’ll attach all references used for your interest/future research.

References

Anthony, T., Copper, C., & Mullen, B. (1992). Cross-racial identification: A social cognitive integration. Personality & Social Psychology Bulletin, 18, 296–301. https://doi.org/10.1177/0146167292183005

Blanton, H., & Mitchell, G. (2011). Reassessing the predictive validity of the IAT II: Reanalysis of heider & skowronski (2007). North American Journal of Psychology, 13(1), 99-106. Retrieved from https://psycnet.apa.org/record/2011-04911-009

Brigham, J. C., Maass, A., Snyder, L. D., & Spaulding, K. (1982). Accuracy of eyewitness  identifications in a field setting. Journal of Personality and Social Psychology, 42, 673–681. https://doi.org/10.1037/0022-3514.42.4.673

Bothwell, R.K., Brigham, J.C., Malpass, R.S. (1989). Cross-racial identification of faces. Personality & Social Psychology Bulletin, 15, 19–25. https://doi.org/10.1177/0146167289151002

Bukach, C.M., Cottle, J., Ubiwa, J., Miller, J. (2012).  Individuation experience predicts other-race effects in holistic processing for both caucasian and black participants. Cognition, 123, 319-324. DOI: 10.1016/j.cognition.2012.02.007

Curby, K. M., Huang, M., & Moer, D. (2019). Multiple paths to holistic processing: Holistic processing of gestalt stimuli do not overlap with holistic face processing in the same manner as do objects of expertise. Attention, Perception and Psychophysics, 81(3), 716-726. http://dx.doi.org.ezproxy.lakeheadu.ca/10.3758/s13414-018-01643-x

Feng, L., Liu, J., Wang, Z., Li, J., Li, L., Ge, L., Tian, J., Lee, K. (2011). The other face of the other race effect: an fmri investigation of the other race categorization advantage. Neuropsychologia, 47, 3739-3749. DOI: 10.1016/j.neuropsychologia.2011.09.031

Fernandez, D. C., Fogerson, P. M., Ospri, L. L., Thomsen, M. B., Layne, R. M., Severin, D., Hattar, S. (2018). Light Affects Mood and Learning through Distinct Retina-Brain Pathways. Cell, 175(1), 71-84.e18. doi:10.1016/j.cell.2018.08.004

Gendron, M., Crivelli, C., & Barrett, L. F. (2018). Universality Reconsidered: Diversity in Making Meaning of Facial Expressions. Current Directions in Psychological Science, 27(4), 211–219. doi:10.1177/0963721417746794

Greenwald, Anthony G., Debbie E. McGhee, and Jordan L. K. Schwartz. 1998. Measuring individual differences in implicit cognition: The Implicit Association Test. Journal of Personality and Social Psychology 74.6: 1464–1480. Retrieved from https://faculty.washington.edu/agg/pdf/Gwald_McGh_Schw_JPSP_1998.OCR.pdf

Hancock, K. J., & Rhodes, G. (2008). Contact, configural coding and the other-race effect in face recognition. British Journal of Psychology, 99(1), 45–56. doi:10.1348/000712607X199981

Hart, A.J., Whalen, P.J., Shin, L.M., McInerney, S.C., Fisher, H., Rauch, S.L. (2000).  Differential response in the human amygdala to racial outgroup vs ingroup face stimuli. NeuroReport, 11, 2351-2355. doi: 10.1097/00001756-200008030-00004

He, Y., Johnson, M. K., Dovidio, J. F., & Mccarthy, G. (2009). The relation between race-related implicit associations and scalp-recorded neural activity evoked by faces from different races. Social Neuroscience, 4(5), 426–442. doi: 10.1080/17470910902949184

Herzmann, G., Willenbockel, V., Tanaka, J.W., Curran, T. (2011). The neural correlates of memory encoding and recognition for own and other race faces. Neuropsychologia, 49,3103-3115. doi: 10.1016/j.neuropsychologia.2011.07.019

Hurtado, E., Haye, A., González, R., Manes, F., Ibáñez, A. (2009). Contextual blending of ingroup/outgroup face stimuli and word valence: LPP modulation and convergence of measures. BMC Neuroscience, 10. doi: 10.1186/1471-2202-10-69

Implicit Association Test (2011). Implicit Association Test [Software]. Retrieved from http://www.implicit.harvard.edu 

Kang, Y., Gray, J. R., & Dovidio, J. F. (2014). The nondiscriminating heart: Lovingkindness meditation training decreases implicit intergroup bias. Journal of Experimental Psychology: General, 143(3), 1306–1313. http://dx.doi.org.ezproxy.lakeheadu.ca/10.1037/a0034150

Kubota, J. T., & Ito, T. A. (2014). The role of expression and race in weapons identification. Emotion, 14(6), 1115–1124. doi: 10.1037/a0038214

Lee, K. M., Lindquist, K. A., & Payne, B. K. (2018). Constructing bias: Conceptualization breaks the link between implicit bias and fear of black americans. Emotion, 18(6), 855-871.  http://dx.doi.org.ezproxy.lakeheadu.ca/10.1037/emo0000347

Malpass, R.S., Kravitz, J. (1969). Recognition for faces of own and other race. Journal of Personality and Social Psychology, 13, 330–334. doi: 10.1037/h0028434

McColl, D., & Nejat, G. (2014). Recognizing Emotional Body Language Displayed by a Human-like 

Social Robot. International Journal of Social Robotics, 6(2), 261–280. https://doi.org/10.1007/s12369-013-0226-7

McCullough, S., & Emmorey, K. (2009). Categorical perception of affective and linguistic facial expressions. Cognition, 110(2), 208-221. doi: 10.1016/j.cognition.2008.11.007

Mendes, W. B., & Koslov, K. (2013). Brittle smiles: Positive biases toward stigmatized and outgroup targets. Journal of Experimental Psychology: General, 142(3), 923–933. doi: 10.1037/a0029663

Mohanty, A., & Sussman, T. J. (2013). Top-down modulation of attention by emotion. Frontiers in Human Neuroscience, 7. doi: 10.3389/fnhum.2013.00102

Richler, J. J., & Gauthier, I. (2014). A meta-analysis and review of holistic face processing. Psychological Bulletin, 140(5), 1281-1302. doi: 10.1037/a0037004

Sagui-Henson S.J. (2017) Cognitive Avoidance. In: Zeigler-Hill V., Shackelford T. (eds) Encyclopedia of Personality and Individual Differences. Springer, Cham. doi: 10.1007/978-3-319-28099-8_964-1

Schiltz, C., Rossion, B. (2006). Faces are represented holistically in the human occipito-temporal cortex. NeuroImage, 32 (3), 1385-1394. doi: 10.1016/j.neuroimage.2006.05.037

Sugden, N. A., & Marquis, A. R. (2017). Meta-analytic review of the development of face  discrimination in infancy: Face race, face gender, infant age, and methodology moderate face discrimination. Psychological Bulletin, 143(11), 1201-1244. http://dx.doi.org.ezproxy.lakeheadu.ca/10.1037/bul0000116

Tanaka, J.W., Kiefer, M., Bukach, C.M. (2004). A holistic account of the own-race effect in face recognition: Evidence from a cross-cultural study. Cognition, 93, B1–B9. doi: 10.1016/j.cognition.2003.09.011

Teles, M. C., Cardoso, S. D., & Oliveira, R. F. (2016). Social plasticity relies on different neuroplasticity mechanisms across the brain social decision-making network in zebrafish. Frontiers in Behavioral Neuroscience, 10, 12. http://dx.doi.org.ezproxy.lakeheadu.ca/10.3389/fnbeh.2016.00016

Wan, L., Crookes, K., Dawel, A., Pidcock, M., Hall, A., & McKone, E. (2017). Face-blind for other-race faces: Individual differences in other-race recognition impairments. Journal of Experimental Psychology: General, 146(1), 102-122. http://dx.doi.org.ezproxy.lakeheadu.ca/10.1037/xge0000249

Wesner, M., Antoniazzi, D., Klein, R., Pollari, A., Cahill, M. (2015). Vigilant and avoidant individuals differentially and holistically discriminate facial emotions, even subliminally. [Abstract]. APS Convention, New York.

Zhou, G., Liu, J., Xiao, N. G., Si, J. W., Li, H., & Kang, L. (2018). The fusiform face area plays a greater role in holistic processing for own-race faces than other-race faces. Frontiers in Human Neuroscience. http://dx.doi.org.ezproxy.lakeheadu.ca/10.3389/fnhum.2018.00220