How do we ‘other’?
Psychological research and practice can enhance or reduce social fairness, and it often does both. This article examines how and why researchers inadvertently reproduce social inequalities when certain habits of thought affect our framing of our research findings.
In 2016 Tamika Cross, a young African-American woman, was a passenger on a flight across the USA when cabin crew asked if any doctor on board could help in an unexpected medical emergency. When Dr Cross came forward, she was initially prevented from helping the ill passenger; the cabin crew took some time to be convinced that she was indeed a doctor. When a white doctor next presented himself, Cross was reportedly told, ‘Thanks for your help, but he can help us.’
Cross’s experience could have been particular, but the viral response to it from other young African-American professional women suggests that it is not. Category norms for professions may fold in not only information about gender, but also race, and their intersections (Cole, 2009). And they can be a matter of life-and-death.
Of course, this was a real-life version of the now famous surgeon riddle:
A man and his son were away for a trip. They were driving along the highway when they had a terrible accident. The man was killed outright but the son was alive, although badly injured. The son was rushed to the hospital and was to have an emergency operation. On entering the operating theatre, the surgeon looked at the boy, and said, ‘I can’t do this operation. This boy is my son.’
In their 2006 study, David Reynolds, Alan Garnham and Jane Oakhill found that about half of people who have not seen this riddle before failed to find the correct answer that the surgeon was the boy’s mother. Surely everyone these days knows that some surgeons are women, so why do they find the riddle difficult? Because the noun surgeon automatically calls to mind imagery of a man.
Such thinking is automatic, in the sense of happening beyond conscious awareness, intent and control. In their 1986 account of ‘norm theory’, Daniel Kahneman and Dale Miller argued that automatically activated working memory representations of categories can render implicit the most prototypical features of the exemplars that ground those category representation (such as the gender of surgeons). People cannot always access the assumption they have made which is limiting their thinking.
Such non-conscious knowledge activation can have such surprising effects on conscious thinking, including the kind of reasoning that psychological scientists do. Recently, inspired by Wason’s (1960) classic experiment on the formation and test of hypotheses, I constructed a ‘celebrity guessing game’, taking the example of the assumed whiteness of film stars as a case in point. Students first saw the names of three film stars and were told that they shared a common attribute that the experimenter held in mind. On each turn of the guessing game, each participant wrote down a hypothesis about what that attribute might be, and generated the name of one new celebrity. The experimenter revealed whether the hypothesis had correctly guessed the true attribute and whether each new celebrity possessed or lacked the attribute on each turn of the game. When the film stars that I provided were black and the attributed-to-be-guessed was race, 90 per cent of participants correctly guessed it, taking about six to seven minutes to win the game. When the film stars were white, a majority of both white and non-white participants failed to discover that all three actors shared a race within the allotted 20-minute period (Hegarty, 2017). As with the gender of surgeons, the race of white film stars was automatically assumed by category norms for such categories as ‘actors,’ ‘Hollywood stars’ or ‘people who go to the Oscars’. The assumption was automatic, but its effects on subsequent conscious reasoning were large.
It is no coincidence that these two attributes folded into category norms that I have discussed so far – being a man and being white – are higher-status identities. Several intellectual traditions of oppressed people have analysed in rich depth how higher-status groups, such as the middle classes (Marx & Engels, 1978), men (de Beauvoir, 1949/2011), white people (DuBois, 1903), or heterosexuals (Warner, 1993) achieve and sustain social power by successfully conflating their interests and perspectives with the universal, impartial or objective definitions.
As professional psychologists, we are not immune to conflating higher-status groups with the norm, and of unthinkingly reinforcing status hierarchies. In the celebrity guessing game studies, participants were more likely to alight on the hypothesis that the three white film stars shared the attribute of ‘race,’ if they had first generated the name of a non-white celebrity and received feedback that this celebrity did not fit the experimenter’s rule. Such thinking would lead a body of researchers to collectively conceptualise race as primarily a characteristic of BAME individuals and only secondarily of white people.
The psychological literature bears the imprint of such thinking. Psychological research articles with the terms culture, race or ethnicity in their title, abstract or keywords have a higher proportion of minorities, and fewer whites, in their samples than does comparable research that does not reference culture, race or ethnicity. Irrespective of their own ethnic self-identification, American psychologists rate an all-white sample more positively for a proposed study addressing personality, but an all-minority sample more positively for a proposed study addressing culture. Psychologists also believe that their fellow researchers would consider the relative importance of personality and cognitive factors to be more explanatory for whites but social and cultural factors more explanatory for ethnic minorities (Causadias et al., 2018).
This cultural mis-attribution bias might particularly sustain inequality because it leads psychologists to conceptualise whites as individuals and ethnic minority individuals through the lens of group stereotypes (Fiske & Neuberg, 1990). There is nothing in our arsenal of tools for establishing reliability or validity that requires us to stand outside of this logic.
The ‘effect to be explained’
The celebrity guessing game and the cultural misattribution bias refer to contexts where psychologists have sampled from only one group on a status hierarchy. However, when differences between groups that differ in status must be explained, the imprint of category norms appears once again. In studies in which students are asked to generate explanations for empirical group difference using their own terms, they write explanations that spontaneously position the higher-status group as the reference group for comparison and the lower-status group as the ‘effect to be explained’. For example, explanations of gender differences in voting preferences would more often reference the claim that ‘women are more liberal’ than the claim that ‘men are more conservative’ (Hegarty, 2006; Miller et al., 1991).
Similarly, about three quarters of the content of explanations of ‘race’ differences and sexual orientation differences reference how and why lower-status groups are different from higher-status groups’ whilst only one quarter of it references how higher-status groups are different from lower-status groups (see Pratto et al., 2008, on ‘race’; and Hegarty & Pratto, 2001, 2004, on sexual orientation). Across these studies, participants explained gender, race or sexual orientation differences in matters as diverse as group differences in voting behaviour, sickness rates, trust in the police, recall of childhood play experiences, responses to medication, drug rehabilitation and charitable donations.
These studies were conducted largely on students, themselves an unexamined ‘normative’ group in psychology research for almost a century (Danziger, 1990). However, published psychological research also attributes empirical gender differences to women and girls more than to men and boys (Hegarty & Buechel, 2006), and some researchers have attributed the development of sexual orientation to homosexual and bisexual groups only, whilst being explicitly aware that such patterns of reasoning constitute a heterosexist bias (see Hegarty & Pratto, 2001, p.732, for evidence; and Herek et al., 1991, on heterosexist research bias).
Such thinking, framing explanation as being disproportionately about lower-status groups, would be an embarrassment to our efforts to keep our science free of social biases on its own. But it can also communicate that lower-status groups are less agentic, less powerful, less worthy, and have a less legitimate claim on power than higher-status groups. The ‘thing’ in the referent position is framed as larger, more important and more immovable than the thing that is figured against it (Gleitman et al., 1996). For example, gender differences in leadership are seen as smaller and less legitimate when women leaders are described as different from men leaders than when men leaders are described as different from women leaders (Bruckmüller, et al., 2012). And some groups – such as left-handers and single people – feel worse about their group identities, showing lowered collective self-esteem (Luhtanen & Crocker, 1992) when they write about how their group differs from its higher-status complement (i.e. right-handers and coupled people) than when asked to write about how such high-status groups differ from their own (Bruckmüller, 2013).
The linguistic framing of explanations, then, not only projects implicit beliefs about group status, it also communicates a set of related beliefs.
Power is ‘done’ not ‘possessed’
These findings pertain to social identity, but this isn’t simply an expression of the basic motive to enhance one’s self-esteem by positioning one’s own group in a good light relative to others. If that were the case, why would women and men, to an equal extent, attribute gender differences to attributes of women rather men (Hegarty, 2006; Hegarty & Buechel, 2006; Miller et al., 1991), or both minority psychologists and white psychologists show the cultural misattribution bias (Causadias et al., 2018). This means that attempts to diversify our samples, or our profession, will not, on their own, simply undo these habits of thinking.
We’re dealing with a discursive form of power; we need to think of power as something that is done rather than possessed (see also Simon & Oakes, 2006). In Foucault’s (1978) terms, psychological science relations of power-knowledge are not static forms of distribution, they are ‘matrices of transformation’.
To install the reverse matrix, psychologists might read cultural psychologist Glenn Adams (2014) on ‘resource extraction’ models of psychological research, observing group differences but failing to address the normative status of economies of excess and self-expansion. We need to take a ‘denaturalising’ turn. Take auditory gaydar: the attempt to detect others’ sexual identities from vocal cues is one kind of ‘gaydar’ judgement that has fascinated some psychologists for two decades (Rule & Alaei, 2016). Gaydar research has recently been critiqued for the essentialist assumptions about sexuality that are re-grounded by its emphasis on finding detectable biologically rooted group differences, particularly differences between gay and straight men (Miller, 2018; Vasilovsky, 2018). Recently, I have worked with Fabio Fasoli and others to begin to reconceptualise auditory gaydar from targets’ perspectives. We asked heterosexual-identified and sexual minority women and men whose voices were being recorded for use in auditory gaydar studies to report whether they believed that their own voice communicated their sexual identity, and whether they desired that it would. Men considered that their voices communicate sexual orientation more than women did, and straight men desired that their voices would communicate their sexual orientation more than did any of the other three groups. Such research findings require a consideration of social status, a matter that is rarely given sustained attention in studies of gaydar accuracy. For gay men, being a target of gaydar risks heterosexist discrimination (see Fasoli et al., 2017). For straight men – the higher-status group – such recognition engenders no such risks.
There are good reasons for social psychologists to quantify the distorting effects of category norms on our thinking about difference. The claim that one’s group has been positioned as ‘the other’ unfairly by psychological research can contribute to discourses of competitive victimhood by which groups engaged in conflict seek to frame themselves as the victim in a current conflict and mutually fuel conflict escalation (Noor et al., 2012). Consequently, it is important to consider the limit conditions of the relationship between status and the framing of explanations described here. Nationality appears to be one such limit. In spite of hierarchies of size and power between these countries neither differences between the Irish and the British, nor differences between the Welsh and the English, are consistently explained by taking the Irish or the Welsh as the ‘effect to be explained’ (Donnelly & Hegarty, 2018; Hegarty, 2013). Despite expectations, the usual asymmetry in explanations was not observed in this context. Experiments can (usefully) prove our hypotheses to be wrong.
Recently, Eitan et al. (2018) extended this line of thinking about ‘the effect to be explained’ to political orientation. These authors presented participants recruited online via social media sites with social psychological conference abstracts presented at the meetings of the Society for Personality and Social Psychology over a 10-year period. The abstracts focused on political topics, and participants evaluated whether the focus of explanation in these articles was skewed toward political conservatives or political liberals (56 per cent vs. 42 per cent of all articles). Interestingly, liberal online participants were more likely than their conservative counterparts to judge the articles as focusing on conservatives. A group of mostly liberal social scientists correctly predicted the direction of this effect but overestimated its magnitude.
That study shows only weak evidence of, as the authors put it, a ‘liberal bias’. Yet they make a strong claim that their report evidences ‘systematic effects of political values on research reports in a scientific field’. I find myself thinking back to that research on race and sexuality. If asked to explain a difference in liberals’ and conservatives’ responses to a blood pressure medication (see Pratto et al., 2008) or a treatment for cancer (see Hegarty & Pratto, 2001), would study participants focus explanations on conservatives more than on liberals by a ratio of 3:1 – like participants focused their explanations on African-Americans and gay men in these earlier studies? Would both liberal and conservative people quickly spot that three conservative celebrities share politics as an attribute but overlook the fact that three liberal celebrities do, mirroring the celebrity guessing game?
I have been slow to call the effects I have found in my experiments ‘bias’, a term that may have an essentially pejorative meaning; I have more often described category norms as creating asymmetries in reasoning. However, I would argue that there’s more evidence of a ‘white bias,’ ‘male bias’ or ‘heterosexual bias’ in psychology than the ‘liberal bias’ that many have discussed in recent years. I hope this article allows psychologists who wish to undo bias in our field in conscious ways to understand a habit that works against our intentions.
- Peter Hegarty is Professor of Social Psychology at the University of Surrey where he leads the research group on Social Emotions and Equality in Relations
Adams, G. (2014). Decolonizing methods: African studies and qualitative research. Journal of Social and Personal Relationships, 31, 467–474.
Bruckmüller, S. (2013). Singled out as the effect to be explained: Implications for collective self-esteem. Personality and Social Psychology Bulletin, 39, 237–249.
Bruckmüller, S., Hegarty, P. & Abele, A.E. (2012). Framing gender differences: Linguistic normativity affects perceptions of power and gender stereotypes. European Journal of Social Psychology, 42, 210–218.
Causadias, J M., Vitriol, J.A. & Atkin, A.L. (2018). Do we overemphasize the role of culture in the behavior of racial/ethnic minorities? Evidence of a cultural (mis)attribution bias in American psychology. American Psychologist, 73, 243–255.
Cole, E.R. (2009). Intersectionality and research in psychology. American Psychologist, 64, 170–180.
Danziger, K. (1990). Constructing the subject: Historical origins of psychological research. Cambridge: Cambridge University Press.
de Beauvoir, S. (2011). The second sex (C. Borde & S. Malovany-Chevallier, Trans.). New York: Vintage. (Original work published 1949)
Donnelly, L. & Hegarty, P. (2018). Explanations of Welsh–English differences. Unpublished data: University of Surrey.
DuBois, W.E.B. (1903). The souls of black folk. Chicago: A.C. McClurg.
Eitan , O., Viganola, D., Inbar, Y. et al. (2018). Is research in social psychology politically biased? Systematic empirical tests and a forecasting survey to address the controversy. Journal of Experimental Social Psychology, 79, 188–199.
Fasoli, F., Hegarty, P., Maass, A. & Antonio, R. (2018). Who wants to sound straight? Sexual majority and minority stereotypes, beliefs and desires about auditory gaydar. Personality and Individual Differences, 130, 59–64.
Fasoli, F., Maass, A., Paladino, M.P. & Sulpizio, S. (2017). Gay- and lesbian-sounding auditory cues elicit stereotyping and discrimination. Archives of Sexual Behavior, 46, 1261–1277.
Fiske, S.T. & Neuberg, S.L. (1990). A continuum of impression formation, from category-based to individuating processes: Influences of information and motivation on attention and interpretation. Advances in Experimental Social Psychology, 23, 1–74.
Foucault, M. (1978). The history of sexuality: Vol. 1. An introduction. New York: Vintage.
Gleitman, L.R., Gleitman, H., Miller, C. & Ostrin, R. (1996). Similar, and similar concepts. Cognition, 58, 321–376.
Hegarty, P. (2006). Undoing androcentric explanations of gender differences: Explaining ‘the effect to be predicted’. Sex Roles, 55, 861–867.
Hegarty, P. (2013). Essential differences: Constructing frames of reference in spontaneous explanations of differences between the British and the Irish. Irish Journal of Psychology, 34, 35–48.
Hegarty, P. (2017). On the failure to notice that white people are white: Generating and testing hypotheses in the celebrity guessing game. Journal of Experimental Psychology: General, 146, 41–62.
Hegarty, P. & Bruckmüller, S. (2013). Asymmetric explanations of group differences: Experimental evidence of Foucault’s disciplinary power. Social and Personality Psychology Compass, 7, 176–186.
Hegarty, P. & Buechel, C. (2006). Androcentric reporting of gender differences in APA journals: 1965–2004. Review of General Psychology, 10, 377–389.
Hegarty, P. & Pratto, F. (2001). The effects of social category norms and stereotypes on explanations for intergroup differences. Journal of Personality and Social Psychology, 80, 723–735.
Hegarty, P. & Pratto, F. (2004). The differences that norms make: Empiricism, social constructionism, and the interpretation of group differences. Sex Roles, 50, 445–453.
Herek, G.M., Kimmel, D.C., Amaro, H. & Melton, G.B. (1991). Avoiding heterosexual bias in psychological research. American Psychologist, 46, 957–963.
Kahneman, D. & Miller, D.T. (1986). Norm theory: Comparing reality to its alternatives. Psychological Review, 93, 136–153.
Luhtanen, R. & Crocker, J. (1992). A collective self-esteem scale: Self-evaluation of one’s social identity. Personality and Social Psychology Bulletin, 18, 302–318.
Lundberg, T., Hegarty, P., & Roen, K. (2018). Making sense of ‘Intersex’ and ‘DSD’: How laypeople with varying experience of intersex/DSD understand and use terminology. Psychology & Sexuality, 9, 161–173.
Marx, K. & Engels, F. (1978). The German ideology, Part I. In R.C. Tucker (Ed.) The Marx–Engels reader (pp.146–200). New York: W.W. Norton.
Miller, A. (2018). Searching for gaydar: Blind spots in the study of sexual orientation perception. Psycholgoy & Sexuality, 9, 188–203.
Miller, D.T., Taylor, B. & Buck, M.L. (1991). Gender gaps: Who needs to be explained? Journal of Personality and Social Psychology, 61, 5–12.
Noor, M., Shnabel, N., Halabi, S. & Nadler, A. (2012). When suffering begets suffering: The psychology of competitive victimhood between adversarial groups in violent conflicts. Personality and Social Psychology Review, 16, 351–374.
Pratto, F., Hegarty, P., & Korchmaros, J. (2008). Who gets stereotyped? How communication practices and category norms lead people to stereotype particular people and groups. In Y. Kashima, K. Fiedler & P. Freytag (Eds.) Stereotype dynamics: Language-based approaches to stereotype formation, maintenance, and change(pp.299–319). Mahwah, NJ: Lawrence Erlbaum.
Reynolds, D.J., Garnham, A. & Oakhill, J. V. (2006). Evidence of immediate activation of gender information from a social role name. Quarterly Journal of Experimental Psychology, 59, 886–903.
Rule, N.O., & Alaei, R. (2016). ‘Gaydar’: The perception of sexual orientation from subtle cues. Current Directions in Psychological Science, 25, 444–448.
Simon, B. & Oakes, P. (2006). Beyond dependence: An identity approach to social power and domination. Human Relations, 59, 105–139.
Vasilovsky, A.T. (2018). Aesthetic as genetic: The epistemological violence of gaydar research. Theory & Psychology, 28, 298–318.
Warner, M. (1993). Introduction. In M. Warner (Ed.) Fear of a queer planet (pp.vii–xxxi). Minneapolis, MN: University of Minnesota Press.
Wason, P.C. (1960). On the failure to eliminate hypotheses in a conceptual task. Quarterly Journal of Experimental Psychology, 12, 129–140.
BPS Members can discuss this article
Already a member? Or Create an account
Not a member? Find out about becoming a member or subscriber