Stereotype Threat: Definition and Examples

Background

The term stereotype threat was first defined by researchers Steele and Aronson as “being at risk of confirming, as self-characteristic, a negative stereotype about one’s group” (Steele et al., 1995).

In other words, stereotype threat refers to an individual’s fear that their actions or behaviors will support negative ideas about a group to which they belong.

For instance, imagine someone worries that performing badly on a test will confirm negative beliefs about the intelligence of their race, gender, culture, or ethnicity. That worry alone is stereotype threat.

The effects are especially evident in the classroom, but they can also follow a person into the workplace and throughout the rest of their lives.

Steele and Aronson’s original study looked at how black and white students performed on a demanding academic test. It used a 30-minute test built from the verbal section of the Graduate Record Examination (GRE). The stakes were high and very real.

Steele and Aronson chose this procedure in response to the racial stereotype that black students are less intelligent or less capable than white students.

The researchers hypothesized that black students primed with the belief that the test was diagnostic of intellectual ability would perform worse than white students.

Un-primed, they expected no such gap. Confirming their suspicions, black participants underperformed white participants when the test was labeled diagnostic of intellectual ability. The pattern was clear. When the same test was labeled non-diagnostic, the two groups performed equally well (Steele et al., 1995).

By labeling the test as diagnostic, the experiment made black students more vulnerable to judgment about their race’s academic ability. Performance suffered as a result. Their mental energy went toward doubt and fear of failure instead of the test.

The finding is often misread as showing that stereotype threat explains the entire Black–White test gap. It does not.

Steele and Aronson’s analyses statistically adjusted for participants’ prior SAT scores, and even in the non-diagnostic condition, black students still scored lower before that adjustment was applied.

What the study demonstrates is narrower but still important: a stereotype’s situational salience can widen an already-existing gap, not create the whole of it (Sackett, Hardison, & Cullen, 2004).

Stereotype Threat Examples

Stereotype threat was first investigated in the context of race and academic performance. Since then, researchers have documented the same pattern across gender, socioeconomic status, and other group identities.

Race and Academic Performance

The original investigation, by Steele and Aronson in 1995, examined the relationship between race and academic performance. The manipulation was simple. Across a series of four experiments, the researchers varied only how a difficult verbal test was described.

When the test was introduced as a genuine measure of intellectual ability, it made a racial stereotype relevant. That relevance vanished when the same test was described as a non-diagnostic exercise.

Later experiments probed the effect from different angles.

One used a word-completion task to check whether the diagnostic framing activated race- and self-doubt-related thoughts. Another found something starker. Merely asking Black students to record their race before an unrelated test was enough to lower their scores.

Threatened students were also more prone to self-handicapping. They reported less sleep and less effort beforehand, and were more likely to avoid race-related preferences. Additional studies have since confirmed the effect on black students’ academic performance more broadly (Osborne, 2001).

Gender and Math Performance

Stereotype threat can also affect performance well beyond race. Spencer and colleagues showed that it may underlie gender differences in advanced math performance (Spencer et al., 1999).

The researchers hypothesized that reducing stereotype threat could help close the math gender gap, based on the cultural belief that women have weaker math abilities. Their findings supported this idea.

When a math test was described as producing gender differences, women performed worse than men. When it was described as not producing gender differences, women performed equally well.

The pattern held. A parallel study using an easier test found no gender gap at all. This fits the theory’s broader claim: threat bites hardest when a task is difficult enough to strain a person’s mental resources. On easier tasks, there is enough spare capacity to absorb the extra load.

Beyond Race and Gender

Stereotype threat has also been extended to socioeconomic status (SES).

Croizet and Claire found that lower-SES students performed worse than higher-SES students on a test described as measuring intellectual ability. That gap disappeared when the same test was described as non-diagnostic (Croizet & Claire, 1998).

These findings challenge the cultural belief that people from lower-SES backgrounds have lesser intellectual ability. That belief is false. Instead, societal stereotypes can hold people back from the achievement they might otherwise attain.

The effect is not limited to stigmatized groups either. Aronson and colleagues tested this directly using highly math-identified white men, a group with no stereotype of mathematical weakness.

Participants were told the study concerned why Asian students tend to outperform white students in mathematics, importing an unfavorable comparison. White men in this threat condition performed significantly worse than those in a control condition (Aronson et al., 1999). The comparison alone was enough.

The result matters because it rules out explanations based on chronic stigma or genuine ability gaps. Whether the identity at stake is race, gender, SES, or something else entirely, stereotype threat is a situational predicament that can befall anyone when a relevant negative comparison becomes salient.

Theories of Stereotype Threat

As these examples show, stereotype threat is a common force that widens racial and gender gaps in performance. What causes the effect?

Recent research connects it to “belonging uncertainty,” the sense that one’s social acceptance and identity are not fully secure (Walton et al., 2007).

The desire for social belonging is a basic human motivation, and members of stigmatized groups may be more uncertain of their social bonds than others.

To protect that sense of belonging, people may go to great lengths to avoid the embarrassment or failure that would come from confirming a negative stereotype about their identity. The stakes feel personal.

A related account comes from Inzlicht, McKay, and Aronson’s (2006) “stigma as ego depletion” theory.

It holds that coping with stigma drains a person’s self-regulatory resources, which then impairs performance on tasks that follow.

Members of a stigmatized group, on this account, have fewer resources left to regulate their actions in a threatening or discriminatory environment. Their capacity starts full, like a fuel tank.

In other words, one’s cognitive abilities can be thought of as a fuel tank that starts on full, but as people face discrimination or negative stereotypes, that fuel is used up by focusing on doubt or concern over their own abilities.

As a result, stigmatized individuals spend so much mental energy worrying about their own talents, skills, or capabilities that they do not have the mental energy left to reach their full potential in following tasks.

Working Memory and the Integrated Process Model

Aim: To explain, mechanistically, how stereotype threat impairs performance, and to unify the many proposed mechanisms into one model (Schmader, Johns, & Forbes, 2008).

Method: Rather than running a new experiment, the authors reviewed the accumulated experimental, cognitive, and physiological evidence. The model centers on working memory.

This is the limited system that holds and manipulates information during difficult reasoning.

Results: The model proposes three processes drain working memory at once. A stress response impairs prefrontal processing. Performance monitoring scans for signs of failure, and effortful suppression fights the anxious thoughts threat provokes. The three compete for the same resource.

Conclusion: The many mechanisms once proposed for stereotype threat share a single bottleneck. They all co-opt working memory, leaving too little capacity for the task at hand (Schmader et al., 2008).

Because all three processes draw on the same limited resource, the net effect is measurably less working memory available for the task itself. Later work supports the model.

Beilock, Rydell, and McConnell (2007) found that threat harms performance most on problems that place heavy demands on working memory. Cadinu, Maass, Rosabianca, and Kiesner (2005) found that women under threat reported more negative task-related thoughts, and that the volume of these thoughts statistically explained their performance decline.

Stereotype Threat and the Achievement Gap

Stereotype threat is especially dangerous because its impacts reach beyond the individual and shape society as a whole. Three areas show this most clearly.

Individual Effects

At the individual level, stereotype threat can increase anxiety and stress as people actively try to disprove negative stereotypes about themselves. Faced with this fear, people often become more disengaged from the subjects or fields the stereotype targets.

By raising the fear of confirming a negative group stereotype, such as lesser intellectual ability, stereotype threat can also produce a lack of confidence, self-defeating behavior, and disengagement.

Ironically, these resulting negative behaviors could cause a self-fulfilling prophecy for the individual who ends up living up to that negative stereotype.

That fear is corrosive. Some people even change their career trajectory to avoid this threat of failure.

A woman interested in math, for instance, might avoid a STEM major for fear of confirming a stereotype about her sex. Multiplied across many individuals, choices like this help explain the low number of women in STEM fields.

Educational Testing

At a societal level, these individual effects combine into a culture in which entire groups are held back. Standardized testing is where this is especially visible.

Exams such as the SAT, ACT, and GRE remain central to college admissions, despite ongoing debate about their fairness. Supporters argue the tests measure academic ability and reasoning. Critics counter that they largely measure access to opportunity.

The racial and socioeconomic gaps in SAT scores support the critics’ view: white and affluent students continue to outperform black, Latinx, and lower-income students.

Because test scores shape access to opportunity and social mobility, these gaps reinforce inequality across generations (Reeves & Halikias, 2017). Stereotype threat is one plausible contributor to why the gap persists.

The situational framing carries a practical lesson. It points to how a testing environment itself is designed: who sits in the room, how the test is described, and whether demographic questions come before or after it.

The lesson is to build identity-safe testing and classroom environments, while being honest that the expected effect sizes are modest.

Workplace Impacts

The effects of stereotype threat do not end at graduation. They can follow people into the workplace, discouraging them from applying for jobs, requesting promotions, or performing with confidence.

Workers who face negative stereotypes about their competence may also show greater anxiety, reduced effort, and less creativity on the job. Over time, this reduces the representation of stigmatized groups in organizations generally.

The pattern repeats in leadership. Women remain underrepresented in senior leadership roles, and ethnic minorities remain underrepresented among CEOs.

Stereotype threat has also been used to argue that standard selection tests understate stigmatized applicants’ true ability. Walton and Spencer (2009) found that grades and test scores can systematically underestimate the latent ability of negatively stereotyped students. This finding has shaped debates about fair hiring and admissions.

The achievement gap, in other words, does not stop at school. It follows people through their careers and the rest of their lives.

How to Fight Stereotype Threat

Given the far-reaching impacts of stereotype threat, researchers have tested several ways to reduce its effects. Four interventions have shown promise.

Teaching About the Threat

Some interventions simply teach people about stereotype threat itself. In one study on women’s math performance, explaining the phenomenon to participants beforehand eliminated the usual gap between men’s and women’s scores (Johns et al., 2005).

The likely reason is re-attribution. Once someone can attribute their test anxiety to a well-documented external phenomenon, rather than to their own inadequacy, the anxiety loses its power to predict failure.

This is an appealingly simple fix. Unlike growth-mindset or self-affirmation programs, it requires no separate exercise or curriculum change. It simply reframes the anxiety a test-taker already feels.

Its effect, like the other interventions here, rests on an assumption: that the underlying stereotype-threat effect is robust to begin with. As the Critical Evaluation section below explains, that assumption does not always hold.

Growth Mindset

A second approach teaches students that intelligence is a trait that can grow, rather than a fixed quantity. In one study, black students encouraged to view intelligence this way reported greater enjoyment and engagement with academics.

The gains were real. They also earned higher grade point averages than students in a control group (Aronson et al., 2002).

The logic is straightforward. If ability can grow, a poor result becomes a temporary setback rather than proof of a fixed group deficit, which takes the sting out of the stereotype.

As with other interventions here, later large-scale attempts found more modest and variable effects than the original study suggested.

The benefit tends to be strongest for lower-achieving and disadvantaged students, which means the effect is real but more conditional than early enthusiasm implied.

Self-Affirmation

A third intervention draws on the same growth-mindset logic but targets self-worth directly. In one study, black students completed a brief in-class writing exercise about values that mattered personally to them, and Cohen and colleagues tracked their grades over the term.

The results were striking.

Students’ grades improved significantly, cutting the racial achievement gap by roughly forty percent (Cohen et al., 2006). Affirming one’s own values, it seems, can build the confidence needed to overcome an internalized stereotype.

A follow-up study went further.

It found the benefits could be recursive: small early gains propagated into better outcomes over a two-year period (Cohen et al., 2009). But self-affirmation is also a leading example of the field’s replication problem.

Later, well-powered replications have often found effects that are small or null. That is a real caveat. The emerging view is that affirmation works, when it works, mainly for specific at-risk students in genuinely identity-threatening settings, not as a universal fix.

Role Models

Finally, role models can play a valuable role in reducing stereotype threat.

The evidence backs this up. One study had college women first read about women who had succeeded in architecture, law, medicine, and invention.

Those women went on to perform significantly better on a difficult mathematics test than women who did not (McIntyre et al., 2003).

Representation does not require in-person exposure to work. Increasing counter-stereotypical role models in television, movies, or literature can shift public perception of stigmatized groups just as effectively.

Early exposure to such role models can also shape children’s aspirations and confidence well into adulthood. Taken together, these interventions give academic institutions and workplaces concrete tools for fighting the threat of stereotypes and building a fairer, less discriminatory environment.

Critical Evaluation

Stereotype threat remains an influential idea, but psychology’s replication crisis has reshaped how researchers view the evidence behind it. Four issues are central to a fair assessment.

1. Publication Bias: the early literature over-represents large effects because small studies with striking results were more likely to be published.
2. Failed Large Replications: the most rigorous, best-powered tests of the effect have often found nothing.
3. Limited Real-World Impact: in high-stakes testing settings, the effect is negligible to small.
4. The Moderator Problem: the theory leans on so many conditions that a null result can always be explained away.

Publication Bias

Aim: To estimate the genuine size of the gender stereotype-threat effect on girls’ maths performance, and to test for publication bias in the evidence (Flore & Wicherts, 2015).

Method: The researchers meta-analyzed 47 effect sizes from published studies on girls’ maths performance under stereotype threat. They checked for publication bias using funnel plots.

Results: The raw estimate showed a small effect in the predicted direction, but small studies reporting large effects were over-represented. Once this bias was corrected for, the effect shrank so much it could no longer be reliably distinguished from zero.

Conclusion: The literature overstates the true effect. The evidence base is considerably weaker than its influence on education policy would suggest (Flore & Wicherts, 2015).

The analysis has real limits. It targets schoolgirls and mathematics specifically, not every population or domain, and bias-correction methods are themselves estimates.

Even so, its central finding, that publication bias inflated the field’s early estimates, has proved robust. Later work echoes the same conclusion.

Failed Large Replications

The most decisive test of the classic gender-maths finding is a registered report by Flore, Mulder, and Wicherts (2018). In a registered report, researchers publish their hypotheses, design, and analysis plan before collecting data, which removes their ability to selectively report favorable results after the fact.

The study tested 2,064 Dutch high-school students, one of the largest and most transparent tests the theory has ever received. The researchers found no overall stereotype-threat effect.

None of the four predicted moderators, domain identification, gender identification, maths anxiety, or test difficulty, made any difference either. Most of the variation in maths performance was explained simply by gender and how strongly a student identified with the subject.

The result was decisive.

A large, transparent, pre-registered study of exactly the population the theory targets returned a clean null result. That does not mean stereotype threat never occurs, but it does mean the founding gender-maths demonstration has not held up under the field’s most rigorous scrutiny.

Limited Real-World Impact

Shewach, Sackett, and Quint (2019) asked a more practical question: how large is stereotype threat in settings that resemble real high-stakes testing?

Their focus was admissions and hiring, not lab studies.

The researchers meta-analyzed stereotype-threat studies and restricted their attention to the subset whose features most closely matched genuine operational testing. In these operational-like settings, the effect ranged from essentially zero to small, and the researchers again detected meaningful publication bias in the underlying literature.

Their conclusion was blunt. Stereotype threat is unlikely to explain much of the real-world gaps in cognitive-test scores it is often invoked to account for.

That does not mean testing environments are irrelevant to fairness. It does mean the size of the fix should be calibrated to a small, not a large, effect.

The Moderator Problem

Stereotype threat depends on several conditions to appear: how invested someone is in the domain, and how strongly they identify with the stereotyped group. It also depends on how difficult the task is, and how subtly the stereotype is cued. Each condition is theoretically reasonable on its own.

Together, though, they create a problem. A null result can always be explained away by pointing to an unmet condition, which makes the theory difficult to falsify and easy to over-fit to whatever the data show.

Stoet and Geary (2012) illustrated the practical cost of this flexibility. They re-examined attempts to replicate the founding gender-maths study.

Its specific crossover pattern, in which the gap appears only when the stereotype is relevant, was rarely reproduced in full. Many so-called replications relied on the same statistical adjustments as the original study, rather than testing the raw, unadjusted effect the popular account describes.

Their assessment was pointed. The strength of the original evidence, they argued, had been overstated relative to the influence it went on to have on education policy.

Taken together, the modern evidence points to a real but modest phenomenon. It is genuine under some laboratory conditions, but smaller, more heterogeneous, and less able to explain real-world achievement gaps than the classic studies implied.

Aronson, J., Fried, C. B., & Good, C. (2002). Reducing the Effects of Stereotype Threat on
African American College Students by Shaping Theories of Intelligence. Journal of Experimental Social Psychology, 38, 113-125.

Cohen, G. L., Garcia, J., Apfel, N. & Master, A. (2006). Reducing the racial achievement gap: A
social-psychological intervention. Science, 313, 1307-1310.

Cohen, G. L., Garcia, J., Purdie-Vaughns, V., Apfel, N., & Brzustoski, P. (2009). Recursive processes in self-affirmation: Intervening to close the minority achievement gap. Science, 324(5925), 400–403. https://doi.org/10.1126/science.1170769

Croizet, J. C., & Claire, T. (1998). Extending the concept of stereotype threat to social class: The
intellectual underperformance of students from low socioeconomic backgrounds. Personality and Social Psychology Bulletin, 24 (6), 588-594.

Flore, P. C., & Wicherts, J. M. (2015). Does stereotype threat influence performance of girls in stereotyped domains? A meta-analysis. Journal of School Psychology, 53(1), 25–44. https://doi.org/10.1016/j.jsp.2014.10.002

Flore, P. C., Mulder, J., & Wicherts, J. M. (2018). The influence of gender stereotype threat on mathematics test scores of Dutch high school students: A registered report. Comprehensive Results in Social Psychology, 3(2), 140–174. https://doi.org/10.1080/23743603.2018.1559647

Inzlicht, M., McKay, L., & Aronson, J. (2006). Stigma as ego depletion: How being the target of
prejudice affects self-control. Psychological Science, 17 (3), 262-269.

Johns, M., Schmader, T., & Martens, A. (2005). Knowing is half the battle: Teaching stereotype
threat as a means of improving women’s math performance. Psychological science, 16 (3), 175-179.

McIntyre, R. B., Paulson, R., & Lord, C. (2003). Alleviating women’s mathematics stereotype
threat through salience of group achievements. Journal of Experimental Social Psychology, 39, 83-90.

Reeves, R. V., & Halikias, D. (2017, August 15). Race gaps in SAT scores highlight inequality and hinder upward mobility. Brookings. Retrieved January 11, 2022, from https://www.brookings.edu/research/race-gaps-in-sat-scores-highlight-inequality-and-hinder-upward-mobility/

Sackett, P. R., Hardison, C. M., & Cullen, M. J. (2004). On interpreting stereotype threat as accounting for African American–White differences on cognitive tests. American Psychologist, 59(1), 7–13. https://doi.org/10.1037/0003-066X.59.1.7

Schmader, T., Johns, M., & Forbes, C. (2008). An integrated process model of stereotype threat effects on performance. Psychological Review, 115(2), 336–356. https://doi.org/10.1037/0033-295X.115.2.336

Shewach, O. R., Sackett, P. R., & Quint, S. (2019). Stereotype threat effects in settings with features likely versus unlikely in operational test settings: A meta-analysis. Journal of Applied Psychology, 104(12), 1514–1534. https://doi.org/10.1037/apl0000420

Spencer, S. J., Steele, C. M., & Quinn, D. M. (1999). Stereotype threat and women’s math
performance. Journal of experimental social psychology, 35 (1), 4-28.

Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of
African Americans. Journal of Personality and social psychology, 69 (5), 797.

Stoet, G., & Geary, D. C. (2012). Can stereotype threat explain the gender gap in mathematics performance and achievement? Review of General Psychology, 16(1), 93–102. https://doi.org/10.1037/a0026617

Osborne, J. W. (2001). Testing stereotype threat: Does anxiety explain race and sex differences
in achievement?. Contemporary educational psychology, 26 (3), 291-310.

Walton, G. M., & Cohen, G. L. (2007). A question of belonging: race, social fit, and
achievement. Journal of Personality and social psychology, 92 (1), 82.

Walton, G. M., & Spencer, S. J. (2009). Latent ability: Grades and test scores systematically underestimate the intellectual ability of negatively stereotyped students. Psychological Science, 20(9), 1132–1139. https://doi.org/10.1111/j.1467-9280.2009.02417.x

Further Information