Stereotyping

Abstract

Stereotyping is a type of attitude: the Medical Subject Headings vocabulary defines it as an oversimplified perception or conception, especially of persons or social groups. In cognitive psychology a stereotype is the belief component of a social category, the traits the mind attaches to a group, and stereotyping is the process of applying those beliefs to an individual. Gordon Allport cast it as the cognitive core of prejudice, ordinary categorization overgeneralized to people. Later work showed that stereotypes can be activated automatically, outside awareness, yet controlled with effort; that their content is organized by warmth and competence; and that merely knowing a stereotype can depress performance. This article covers what stereotyping is, how stereotypes form, their automatic and controlled operation, their content, and their consequences, with a worked example and three demonstrations.

Keywords: stereotyping, social categorization, implicit stereotype, stereotype content, stereotype threat

A stereotype is what the mind expects of a person once it has sorted them into a group. Told only that someone is an engineer, a teenager, or a nurse, a perceiver immediately has a set of expectations about their traits, and those expectations guide attention, memory, and judgment before any individuating information arrives. This is efficient, because categorization is how the mind copes with more social information than it could ever process case by case, and it is also the root of the problem, because the category is applied to a particular person who may fit it poorly or not at all. Stereotyping sits at exactly this junction between a useful cognitive shortcut and an unfair prejudgment.

This article treats stereotyping as a measurable social-cognitive process rather than a moral failing alone. It defines the stereotype as the belief component of a group category, traces how such beliefs form from ordinary mechanisms of categorization and covariation detection, sets out the distinction between automatic activation and controlled application that organizes the modern field, describes how the content of stereotypes falls along the dimensions of warmth and competence, and examines the consequences of stereotyping for perception, action, and the performance of the people stereotyped. The organizing claim is that stereotyping is normal cognition applied to social groups, which is precisely why it is so persistent and so consequential.

What Stereotyping Is

A stereotype is a generalization about the attributes of a social group that is applied to its individual members. The Medical Subject Headings vocabulary defines stereotyping as an oversimplified perception or conception, especially of persons or social groups, and files it as a type of attitude, reflecting that a stereotype is a stored evaluative belief rather than a passing impression. Psychologically, the stereotype is the cognitive component of a broader orientation toward a group whose two other components are prejudice, the affective or evaluative response, and discrimination, the behavioral one. Stereotyping is the cognitive act: sorting a person into a category and importing the category's associated traits as expectations about that person.

The foundational treatment is Gordon Allport's, which located stereotyping within normal thought rather than apart from it. Allport argued that the human mind must think with the aid of categories, that categories form large classes and clusters that guide daily adjustment, and that prejudice is in large part an overextension of this ordinary process to human groups (Allport, 1954). On this account a stereotype is not a special pathology but a category applied where it fits badly, which is why stereotyping is universal and resistant to correction: the same machinery that lets a perceiver recognize a chair on sight lets them assign traits to a person on the basis of group membership. Later reviews retained this framing while making the construct precise, treating stereotypes as knowledge structures that influence the encoding, storage, and retrieval of information about people, with both functional benefits and systematic costs (Hilton & von Hippel, 1996).

How Stereotypes Form

Stereotypes arise from the mind's ordinary tools for finding structure in experience, and two of those tools are central. The first is social categorization: the perceiver divides people into groups, and the mere act of categorizing exaggerates the similarities within a group and the differences between groups, so that members come to seem more alike, and more unlike outsiders, than they are. The division is rarely neutral. A perceiver separates their own ingroup, the group to which they belong, from an outgroup to which they do not, and outgroup members are seen as more interchangeable than ingroup members, a homogeneity effect that makes a single stereotype feel apt for every outsider. Categorization is cognitively efficient, freeing resources for other tasks, which is why it is engaged so readily and so hard to suppress (Macrae & Bodenhausen, 2000). Once a group exists for the perceiver, a stereotype can attach to it.

The second tool is covariation detection, and here a well-documented error shows how a stereotype can form with no basis in fact. In the illusory correlation, people perceive a relationship between two variables that are in truth unrelated, because a co-occurrence of two distinctive events is especially memorable. David Hamilton and Robert Gifford presented participants with statements pairing members of a larger group and a smaller group with desirable and undesirable behaviors, keeping the ratio of desirable to undesirable behavior identical across the two groups, so that there was no actual association between group and behavior (Hamilton & Gifford, 1976). Because membership in the smaller group was rarer and undesirable behavior was rarer, their co-occurrence was doubly distinctive, drew extra attention, and was over-remembered; participants ended up believing the minority group behaved worse. A negative stereotype of a minority had been manufactured purely from the distinctiveness of a statistical pairing. The first demonstration lets the reader reproduce this effect from the raw frequencies.

Automatic Activation and Controlled Application

The insight that organizes the modern field is that having a stereotype and using it are different events. Patricia Devine drew the distinction sharply, proposing that a well-learned cultural stereotype is activated automatically whenever its target group is encountered, equally in people high and low in prejudice, because both have learned the same cultural associations; what differs is the second step, in which low-prejudice people are motivated to inhibit the activated stereotype and replace it with their egalitarian beliefs, an effortful and controllable process (Devine, 1989). This dissociation reframed prejudice reduction. If activation is automatic but application is controlled, then the stereotype coming to mind is not itself the failure; the failure is letting it govern judgment, and that step is breakable, a habit that can be unlearned with sustained effort.

Later work specified when the controlled step succeeds or fails. Ziva Kunda and Steven Spencer argued that activation and application are each governed by the perceiver's goals: a stereotype is more likely to be activated when it serves a current goal such as self-enhancement or comprehension, and more likely to be applied when the perceiver lacks the motivation or cognitive resources to set it aside (Kunda & Spencer, 2003). Cognitive load matters because application can be the path of least resistance. Macrae and Bodenhausen showed that stereotyping is in part a resource-conserving strategy: under time pressure or competing demands, perceivers lean more heavily on category-based expectations because individuation is costly (Macrae & Bodenhausen, 2000). The second demonstration separates the two stages, letting the reader see a stereotype activate and then watch controlled processing either apply or inhibit it.

The Content of Stereotypes: Warmth and Competence

Stereotypes differ not only in strength but in content, and that content proves to be highly structured. The stereotype content model of Susan Fiske and colleagues holds that group stereotypes are organized by two fundamental dimensions, warmth and competence, and that a group's position on them is predicted by two features of its relationship to the perceiver's group: perceived competition predicts low warmth, and perceived status predicts high competence (Fiske et al., 2002). The two dimensions yield four quadrants, and crucially many groups fall into the mixed, off-diagonal cells. A group seen as warm but incompetent, such as a stereotype of the elderly, draws pity; a group seen as competent but cold draws envy; the combinations map onto distinct emotional prejudices rather than a single like or dislike.

The warmth and competence framework has been refined and challenged. Alex Koch and colleagues proposed an alternative two-dimensional space recovered directly from how people spontaneously sort groups, labeling the axes agency or socioeconomic success, conservative versus progressive beliefs, and communion, the ABC model (Koch et al., 2016). The debate turns on which dimensions are primary and whether a belief dimension is needed alongside the warmth and competence pair, but both frameworks agree on the deeper point that stereotype content is low-dimensional and systematic: a small number of axes capture most of how social groups are perceived. Implicit measures reinforce this, showing that the associations driving stereotype content often operate below the level of deliberate report (Greenwald & Banaji, 1995). The third demonstration places groups on the warmth and competence plane and reads off the predicted emotional response.

Consequences: From Split-Second Judgments to Performance

Because stereotypes can be activated automatically and applied under load, their consequences reach into decisions made too fast for deliberation. Keith Payne showed that priming participants with a Black face led them to identify a tool as a weapon more quickly, and, when responses were forced to be fast, to misidentify harmless objects as weapons more often, a demonstration that a racial stereotype can bias the perception of a threatening object itself (Payne, 2001). Joshua Correll and colleagues extended the finding to the decision to shoot: in a videogame task, participants were faster to shoot armed Black targets and to decline to shoot unarmed White targets, and under time pressure made more errors in the stereotype-consistent direction, the shooter bias (Correll et al., 2002). These effects appear even in people who disavow prejudice, because they trade on automatically activated associations rather than endorsed beliefs (Greenwald & Banaji, 1995).

Stereotyping also harms the people it targets through their own awareness of it. Claude Steele and Joshua Aronson demonstrated stereotype threat: when a negative stereotype about one's group is made salient in a performance setting, the fear of confirming it consumes working memory and depresses performance, independent of any actual ability difference (Steele & Aronson, 1995). Black students performed worse on a difficult test when it was described as diagnostic of ability, framing that invoked a racial stereotype, than when the identical test was described as nondiagnostic. The stereotype did its damage without anyone else having to endorse or apply it; the target's knowledge that the stereotype existed was enough. Stereotyping is therefore consequential on both sides of the social encounter, shaping how a person is seen and, through threat, how they can perform.

Worked Example: Manufacturing an Illusory Correlation

Consider the illusory correlation paradigm with concrete frequencies. Participants read short statements, each pairing a member of Group A, the larger group, or Group B, the smaller group, with a desirable or an undesirable behavior. The key feature is that the ratio of desirable to undesirable behavior is identical in the two groups, so there is no real association between group and conduct.

- Group A (majority): 18 desirable and 9 undesirable behaviors, 27 statements in all. - Group B (minority): 8 desirable and 4 undesirable behaviors, 12 statements in all.

The contingency is exactly zero. Group A's undesirable share is `9 / 27 = 0.333`, and Group B's is `4 / 12 = 0.333`: the two are identical, so a person's group membership carries no information about the odds of undesirable behavior. Both groups follow the same `2 : 1` ratio of desirable to undesirable conduct. Yet the raw counts make one combination distinctive. Across all `27 + 12 = 39` statements, undesirable behaviors are rare (`13` of `39`) and minority members are rare (`12` of `39`), so the pairing of the two, a minority member behaving undesirably, is the rarest cell at just `4` of `39` observations. Its double rarity makes it vivid and over-encoded.

When participants later estimate how often each group behaved undesirably, the distinctive cell is inflated. A typical result has them attributing roughly `6` undesirable behaviors to Group B rather than the actual `4`, a perceived undesirable share of `6 / 12 = 0.500` against the true `0.333`, an overestimate of `0.500 − 0.333 = 0.167`. The same participants stay close to accurate for the majority group, whose behaviors were never distinctive. From data containing no real association, a negative stereotype of the minority has been produced, and it was produced by a general memory bias toward distinctive co-occurrences, not by any prior animus (Hamilton & Gifford, 1976). This is why Allport's claim that stereotyping grows from ordinary cognition is so important: the bias needs no motive to get started (Allport, 1954).

Discussion

The study of stereotyping is the study of how normal cognition, applied to social groups, produces systematically unfair judgments. The field's arc runs from Allport's insight that stereotyping is categorization overextended to people, through the demonstration that stereotypes can form from pure statistical distinctiveness, to the organizing distinction between automatic activation and controlled application, and on to a structured account of what stereotypes contain and what they cost (Allport, 1954; Hamilton & Gifford, 1976; Devine, 1989; Fiske et al., 2002). The decisive move was dissociating the having of a stereotype from its use, because it located the point at which intervention is possible: not in preventing a cultural association from being learned, but in strengthening the controlled processing that decides whether the association governs behavior (Devine, 1989; Kunda & Spencer, 2003).

What the mature field shows is that stereotyping is both cognitively ordinary and socially corrosive, and that these facts are connected rather than opposed. The same efficiency that makes category-based expectations adaptive makes them intrude under load and operate below awareness (Macrae & Bodenhausen, 2000; Greenwald & Banaji, 1995). Part of why they resist correction is subtyping: a member who plainly disconfirms the stereotype is quarantined into a special subcategory, the exception that preserves the rule, so the general belief survives the counterevidence that ought to overturn it. The open problems concern how far the costs can be reduced without sacrificing the efficiency, how stable implicit stereotypes are over time, and how the competing content models relate. Debates over the reliability and predictive validity of implicit measures remain active, and they matter because much of the evidence that stereotyping operates automatically rests on those measures. The construct is unusually well integrated across its levels, from the millisecond dynamics of a weapon-identification task to the decade-scale change in a society's attitudes, but the bridges between those levels are still being built.

Cognitive Implications

Stereotyping is a model case of how a general-purpose cognitive mechanism produces a specific social outcome. Categorization, covariation detection, and the automatic retrieval of well-learned associations are all domain-general processes the mind uses everywhere; stereotyping is what happens when they are turned on human groups. This is why the phenomenon cannot be dissolved simply by exhorting people to think better, and why it recurs across cultures and eras: it is a side effect of cognition working as designed (Allport, 1954; Macrae & Bodenhausen, 2000).

The activation and application distinction also carries a general lesson for the study of the mind: that a single observable behavior can issue from more than one underlying process, and that progress comes from prying those processes apart. The weapon-identification and shooter tasks make this concrete, using response speed and error patterns to separate an automatic bias from the controlled correction that may or may not override it (Payne, 2001; Correll et al., 2002). Stereotype threat adds a further twist, showing that the mere cognitive representation of a stereotype, held by its target, can consume working memory and alter performance (Steele & Aronson, 1995). In each case the lesson is that the social consequence is downstream of a specifiable cognitive mechanism, which is what makes stereotyping a cognitive-psychological problem and not only a social one.

Current Directions

The most active recent work asks how the content and the trajectory of stereotypes can be measured at scale. Alex Koch and colleagues recovered the dimensions of stereotype content directly from how people spontaneously sort large sets of groups, arguing that agency or socioeconomic success and a conservative-progressive belief dimension, alongside communion, organize the space, and sharpening the debate with the warmth and competence account over which dimensions are primary (Koch et al., 2016). The disagreement is productive because both camps now test their models against the same spontaneous-sorting and large-group data rather than against a fixed list of traits.

A second direction tracks stereotypes as they change. Naomi Ellemers synthesized the evidence on gender stereotypes, documenting how resilient they are, how they operate through descriptive and prescriptive expectations, and how they continue to shape behavior even as explicit endorsement declines (Ellemers, 2018). Complementing this, Tessa Charlesworth and Mahzarin Banaji used a large repository of implicit and explicit measures to chart long-term change, finding that some implicit attitudes, such as those toward sexual orientation, moved substantially toward neutrality over roughly a decade while others barely shifted (Charlesworth & Banaji, 2019). The finding that implicit associations, long treated as sticky, can move at the level of a whole society reframes a central question of the field, from whether implicit stereotypes change to which ones change, how fast, and why.

Common Misconceptions

A stereotype is the same thing as prejudice.

No. The stereotype is the cognitive component, the belief about a group's attributes; prejudice is the affective component, the evaluative feeling; discrimination is the behavioral component. A person can hold a stereotype in memory without endorsing it or acting on it (Allport, 1954; Devine, 1989).

Only prejudiced people have stereotypes in mind.

No. Well-learned cultural stereotypes are activated automatically in low-prejudice and high-prejudice people alike, because both have learned the same associations. The difference lies in the controlled step of whether the activated stereotype is inhibited or applied (Devine, 1989; Kunda & Spencer, 2003).

Stereotypes only form when there is some real group difference.

No. An illusory correlation shows that a stereotype can form from data containing no real association at all, purely because a distinctive pairing of a rare group and a rare behavior is over-remembered (Hamilton & Gifford, 1976).

Stereotypes only harm their targets when others act on them.

No. Stereotype threat shows that a person's own awareness of a negative stereotype about their group can depress their performance, by consuming working memory, even when no one else applies the stereotype to them (Steele & Aronson, 1995).

Glossary

Attitude.

A stable evaluative orientation toward an object or class of objects; in the Medical Subject Headings vocabulary stereotyping is classified as a type of attitude.

Automatic stereotype activation.

The involuntary triggering of a group's associated traits on encountering the group or a cue to it, occurring without intention and often outside awareness.

Controlled processing.

Effortful, intentional cognition that can inhibit an activated stereotype and substitute an individuated or egalitarian judgment, provided motivation and resources are available.

Discrimination.

The behavioral component of a group orientation; differential treatment of a person on the basis of group membership, as distinct from the stereotype belief and the prejudice feeling.

Illusory correlation.

The perception of a relationship between two variables that are in fact unrelated, arising when a distinctive co-occurrence is over-encoded; a mechanism by which stereotypes form without any real group difference.

Implicit Association Test.

A reaction-time measure that infers the strength of automatic associations between a group and attributes from the ease of pairing them, widely used to index implicit stereotypes.

Implicit stereotype.

A group-trait association that operates below deliberate awareness or control and can influence judgment even in people who explicitly reject it.

Ingroup.

A social group to which a person belongs; categorization tends to favor the ingroup and sharpen its contrast with outgroups.

Outgroup.

A social group to which a person does not belong; outgroup members are more readily seen as interchangeable and are more often the object of stereotyping.

Prejudice.

The affective or evaluative component of a group orientation; a positive or negative feeling toward a group, of which the stereotype is the cognitive counterpart.

Shooter bias.

The tendency, under time pressure, to decide to shoot an armed target more readily and to make more stereotype-consistent errors when the target belongs to a stereotyped group.

Social categorization.

The sorting of people into groups, which exaggerates within-group similarity and between-group difference and is the cognitive precondition for stereotyping.

Stereotype content model.

Fiske and colleagues' account holding that group stereotypes are organized by warmth and competence, predicted by perceived competition and status and mapping onto distinct emotional prejudices.

Stereotype threat.

The depression of performance that follows from a person's awareness of a negative stereotype about their group in a relevant setting, independent of any ability difference.

Stereotype.

A generalization about the attributes of a social group that is applied to its individual members; the cognitive component of a group orientation.

Subtyping.

The mental quarantining of a group member who disconfirms a stereotype into a special subcategory, which protects the overall stereotype from revision.

Key Researchers

Gordon W. Allport

(Harvard University; 1897–1967). Personality and social psychologist whose book The Nature of Prejudice framed stereotypes as the cognitive component of prejudice and located stereotyping in the ordinary process of categorization. Wikipedia · Wikidata

Mahzarin R. Banaji

(Harvard University). Social psychologist whose work on implicit social cognition and the Implicit Association Test demonstrated that stereotypes operate outside conscious awareness and control. Faculty · Wikipedia

Patricia G. Devine

(University of Wisconsin–Madison). Social psychologist whose dissociation model distinguished automatic stereotype activation from controlled inhibition, reframing prejudice reduction as a breakable habit. Faculty · Wikipedia

Susan T. Fiske

(Princeton University). Social psychologist whose stereotype content model maps group stereotypes onto warmth and competence and predicts distinct emotional prejudices from perceived status and competition. Faculty · Wikipedia

Anthony G. Greenwald

(University of Washington). Social psychologist who, with Banaji, formalized implicit social cognition and invented the Implicit Association Test, grounding the measurement of unconscious stereotypes. Faculty · Wikipedia

Claude M. Steele

(Stanford University). Social psychologist whose stereotype threat research showed that awareness of a negative group stereotype can itself depress performance, linking stereotyping to real achievement gaps. Faculty · Wikipedia

Henri Tajfel

(University of Bristol; 1919–1982). Social psychologist whose work on social categorization and the minimal group paradigm established how mere group membership drives intergroup differentiation and stereotyping. Wikipedia · Wikidata

Frequently Asked Questions

What is a stereotype?

A stereotype is a generalization about the attributes of a social group that is applied to its individual members. It is the cognitive component of a group orientation, the belief about what the group is like, which the mind imports as an expectation once a person has been sorted into the group.

What is the difference between a stereotype, prejudice, and discrimination?

They are the three components of an orientation toward a group. The stereotype is the belief or cognitive component, prejudice is the feeling or affective component, and discrimination is the behavior. A person can hold a stereotype without endorsing the prejudice or acting on it (Devine, 1989).

How do stereotypes form?

They form from ordinary cognition: social categorization sorts people into groups and exaggerates group differences, and covariation detection can manufacture a group-trait association even from unrelated data, as in the illusory correlation (Hamilton & Gifford, 1976; Macrae & Bodenhausen, 2000).

Are stereotypes automatic, or can they be controlled?

Both. A well-learned stereotype is activated automatically on encountering its target group, but whether it then governs judgment depends on controlled processing, which can inhibit it when the perceiver has the motivation and the cognitive resources to do so (Devine, 1989; Kunda & Spencer, 2003).

What is the stereotype content model?

It is the account that group stereotypes are organized by two dimensions, warmth and competence, with a group's position predicted by perceived competition and status. The combinations map onto distinct emotional prejudices such as pity, envy, admiration, and contempt (Fiske et al., 2002).

What is stereotype threat?

Stereotype threat is the drop in performance that follows when a person is made aware of a negative stereotype about their group in a setting where it applies. The worry about confirming the stereotype consumes working memory and impairs performance, independent of actual ability (Steele & Aronson, 1995).

Are stereotypes accurate?

Some capture real average differences while many do not, but accuracy is not the core issue. Even a statistically defensible generalization misleads when applied to a particular person, and stereotypes can form with no basis in fact and then resist disconfirmation through subtyping (Allport, 1954; Hamilton & Gifford, 1976).

Can stereotypes change?

Yes, including implicit ones once thought fixed. Large-scale data show that some implicit attitudes moved substantially toward neutrality over about a decade while others changed little, so the question has shifted from whether implicit stereotypes change to which ones change and why (Charlesworth & Banaji, 2019; Ellemers, 2018).

References