Membership in a stereotyped group is the one condition that must always be present in order for stereotype threat to occur. Everyone is a member of at least one stereotyped group, and is vulnerable to being affected by stereotype threat in different situations. Although greatly contested, women and members of ethnic minorities are not the only groups affected by stereotype threat. Research on stereotype threat has shown that a diverse set of groups can be affected by stereotype threat. Aronson, Lustina, Good, Keough, Steele, and Brown (1999) showed that White males performed worse on a math test when the performance of Asians in mathematics was made salient. In this context, White males are part of the stereotyped group because of the stereotype that White males are worse at math than Asian males.
Individuals Who Identify Strongly With Groups
Certain individuals are more vulnerable to stereotype threat than others because their memberships are salient to them in many situations. These individuals strongly identify with groups which use gender or ethnicity as inclusive criteria. This strong identification with groups can make stereotypes chronically accessible (what does this mean?). In turn, they will increase stereotype threat more often and in varying situations. Whereas certain individuals need to have their stereotyped group membership made salient, these individuals are always aware of their stereotypes. Marx, Stapel and Muller (2005) found that the performance of women on a math test was inversely proportionate to the degree to which their gender was emphasized. Cole, Matheson and Anisman (2007) found similar results. Higher ethnic identification predicted greater psychological distress and poorer performance for minority students in their first year of college.
Stereotype Belief and Knowledge
Though stereotype threat can arise in situations where the person is unconsciously aware of the stereotype, it is more easily processed when the person is aware or knowledgeable of the stereotype. Since adults are usually aware of many stereotypes, McKown and Weinstein (2003) conducted a study that examined the knowledge of stereotypes by children and its effects on their performance in domains that produce stereotype threat in adults. The researchers found that children who were aware of stereotypes were more likely to show effects of stereotype threat. Belief of stereotypes is another factor that affects the degree to which a stereotype threat is experienced. In a study by Schmader, Johns, and Barquissau (2004), women who endorsed gender stereotypes about women’s math ability did worse on a task in the stereotyped domain when under stereotype threat more so than their counterparts who did not endorse a gender stereotype.
Consequences of Stereotype Threat
Decreased Performance on Tasks
Test performance in academic settings is the context in which most stereotype threat research is conducted. Cole, Matheson and Anisman (2007) found that ethnic minority students at a predominantly White Canadian institution showed increased levels of anxiety and depression compared to the White students, and had lower grades at the end of the school year. Scores on Raven’s Advanced Progressive Matrices (APM) task, a task that has been described as being a pure measure of cognitive ability and culture-free, have been shown to be affected by stereotype threat (Brown & Day, 2006). Performance on tasks in other domains has also been affected by stereotype threat. Undergraduate female golfers performed more poorly when the stereotype of females being poor golfers was purported either subtly or blatantly (Stone McWhinnie, 2008). The reason for reduced performance when under stereotype threat has been attributed to factors such as anxiety (Max & Stropel, 2006), physiological arousal (Blascovich et al., 2001) and reduced working memory capacity (Schmader and Johns, 2003).
Changes in Appraisal of Situations
When faced with stereotype threat, individuals use different strategies to account for lapses in their performance. Some individuals attribute internal reasons/causes as to why they failed the task. In a study by Koch, Müller, and Sieverding (2008), participants were asked to find a document on a computer and save it to an external storage device. Participants were told that women perform worse on the task than men (stereotype threat condition for women) or that women perform better on the task than men (stereotype threat condition for men). Women in the stereotype threat condition were more likely to make internal attributions for their failures than men in the same experiment. Self-handicapping is another strategy used by individuals affected by stereotype threat. These individuals erect barriers to performance in order to provide attributions for their failure (Stone, 2002). A specific type of self handicapping is task discounting. Tasks are critiqued in some way in order to explain poor performance. A study by Lesko and Corpus (2006) found that women who were taking part in a math task and under stereotype threat were more likely to agree with statements such as “this test is not an accurate measurement of my math ability”.
Reducing Stereotype Threat
Reframing the Task
Stereotype threat can arise when task descriptions or test instructions make threatened identities and negative stereotypes more accessible. Ergo, stereotype threat can be reduced by reframing a task in such a way that stereotypes are not invoked or made salient. Steele and Aronson (1995) were able to reduce stereotype threat by explicitly stating that a test was not diagnostic in nature. Though this strategy is unrealistic in regular testing situations, other task reframing methods have been shown to be equally effective. Gender stereotypes are common, and affect the performance of women in stereotyped domains (Spencer, Steele & Quinn, 1999). However, gender stereotypes and their subsequent effect on test performance can be reduced by reframing the task. For example, Spencer, Steele and Quinn (1999) showed that when both male and female participants completed a mathematics test and were told that there would be gender differences, the males outperformed the females. Contrarily, when the task was reframed and the participants were told that there would be no gender differences, the scores of the male and female participants were equivalent. Another task reframing strategy is addressing the fairness of the task. Good, Aronson and Harder (2008) were able to reduce stereotype threat present on a mathematical ability test by assuring the test takers that “this mathematics test has not shown any gender differences in performance or mathematics ability”. Even the though the tests outlined in the above studies were diagnostic in nature, reframing the tasks was able to negate potential stereotype threat effects. Task reframing strategies can be employed by testing agencies in order to reduce the potential effects of stereotype threat, which can affect the scores of test takers belonging to certain demographics on standardized tests.
De-emphasizing Threatened Identities
Stereotype threat can be reduced in situations where the threatened social identities are made less salient. Threatened social identities can consist of gender, ethnicity, socioeconomic status and other domains that are often stereotyped. Stricker and Ward (2004) conducted field studies that examined the effects of asking about ethnicity prior to Advanced Placement (AP) calculus exams. A re-analysis of their data by Danaher and Crandall (2008) found that soliciting identity information at the end of the exam decreased sex differences in test scores by 33% compared to when identity questions were asked at the beginning of the exam. Danaher and Crandall concluded that if the demographic questions were asked at the end of AP calculus exams, 4700 additional female students would receive a credit annually.
Stereotype threat can also be reduced if individuals are encouraged to think in ways that reduce the significance of the threatened identity. Ambady, Paik, Steele, Owen-Smith and Mitchell (2004) found that when women were individuated by having aspects of the self made more salient, they performed better on a difficult math test than women who were not individuated. The researchers concluded that individuation allows individuals to distance themselves from the threatened social identity. Emphasizing the similarities between groups has also been shown to reduce stereotype threat. Rosenthal, Crisp and Sue (2007) found that generating similarities between men and women in academic settings led to higher performance expectations and improved performance by women in a stereotyped domain. Rosenthal, Crisp and Sue (2007) selected mathematical ability as the stereotyped domain in which similarities between ingroup and outgroup employment strategies would be tested.
Individuals with multiple social identities have a low risk of being affected by stereotype threat. Gresky, Ten Eyck, Lord and McIntyre (2005) examined the role of multiple social identities in undergraduate university students. The participants were instructed to answer questions from what was supposedly a new version of the GRE math exam. Stereotype treat was induced by informing the participants that men usually outperform women on math tests. The participants were then split into three conditions. Before the math test, some subjects were instructed to create a simple self-concept map whereas others were instructed to create a complex self-concept map. Subjects in the control condition did not create a self-concept map. The female participants who did not create a self-concept map or created a simple self-concept map performed poorly on the math test whereas the female participants who were instructed to create a complex self-concept map were unaffected by the stereotype threat manipulation. Creating self-concept maps did not have an effect on the scores of the male participants, presumably because they were not affected by the stereotype threat manipulation. Consequently, encouraging stereotype threatened individuals to think of themselves as multifaceted individuals reduces vulnerability to stereotype threat effects.
Though all people have multiple identities, some are more vulnerable to stereotype threat than others. This can be explained by the fact that different social identities are highlighted to different degrees in individuals. Having stereotyped social identities that are easily highlighted (not the right word) increases the vulnerability of an individual to stereotype threat. Mcglone and Aronson (2006) demonstrated this effect by instructing participants to complete the Vandenberg Mental Rotation Test (VMRT) in varying conditions of identity salience. This salience was achieved by having participants complete questionnaires that highlighted different social identities. For the male participants, performance on the VMRT was best when gender was made salient and worst when their status as citizens of the Northeast was made salient. The female participants performed the best when their college identity was made salient and performed the worst when their gender was emphasized.
Stereotype threat acts on an individual’s fear of confirming negative stereotypes (Steele & Aronson, 1995). Self-affirmation can be used as a strategy to protect the self from perceived threats. Self-worth can be increased by encouraging people to think about characteristics, roles and skills that they are proud of or view as important. Schimel, Arndt, Banko and Cook (2004) examined whether self-affirmation and affirming self-worth can protect individuals from stereotype threat. Before completing a math test, female participants were randomly assigned sentence fragments to complete that provided either intrinsic affirmation, extrinsic affirmation or no affirmation. The participants were also placed in either a stereotype threat or control condition. The researchers found that women in the stereotype threat condition who reaffirmed their self worth by completing the sentence fragments that provided intrinsic motivation performed better on the exam in the stereotype threat condition than in the control condition. Conversely, women in the stereotype threat condition who extrinsically self-affirmed performed worse in the stereotype threat condition compared to the control condition. Cohen, Garcia, Apfel and Master (2006) came to similar conclusions when studying seventh grade students in racially diverse middle schools. Students were assigned to self-affirmation conditions and no self-affirmation conditions. The students in the self-affirmation condition were instructed to write an essay that indicated values that were important to them and why they were important to them. Students who did not self-affirm were instructed to write an essay about their least important values and why they were not important. The researchers found that African-American students in the self-affirmation condition performed 0.3 grade points better over the semester compared to African-American students who did not self-affirm. These results are notable because African-American students are the demographic that is usually most affected by stereotype threat in academic settings. As a result, enhancing the individual’s internal coping capacity through providing affirmation has shown to improve the performance of individuals that are most susceptible to stereotype threat.
Providing Role Models
Individuals who focus on outgroup members who perform well on a specific domain have a negative effect on their performance on tasks from that domain. Huguet and Régner (2007) showed that the performance of female participants on a math test in a mixed environment was negatively affected by their thoughts about males who performed well in mathematics. This type of stereotype threat effect can be reduced by providing ingroup role models that are proficient in a specific domain. Marx and Roman (2002) examined this effect by providing females with role models who were competent in math. Male and female participants were given a difficult math test to complete. When the experimenter was a male, the female participants performed poorer on the exam than the male participants. However, when the test was administered by a female who embodied subject confidence, the female participants performed equally as well as the male participants.
The role models provided do not need to be present in order to reduce stereotype threat. Various studies have shown that stereotype threat can be reduced by providing role models through priming. McIntyre et al. (2005) examined the relation between exposure to positive role models through priming and reduction of stereotype threat effects. Undergraduate students were asked to participate in two studies, and were told that women perform worse than men on math tests. The researchers also manipulated the tasks by having the participants read anywhere from 0-4 essays describing successful women. The researchers found that the performance of the female participants increased as the number of essay read increased, supporting the finding that providing role models decreases stereotype threat effect.
Attributing Difficulties to External Factors
The effects of stereotype threat can be reduced if difficulties on tasks are attributed to external factors rather than the self. Ben-Zeev, Fein, and Inzlicht (2005) illustrated the effect of attributing difficulties to external factors on stereotype threat. Women who identified highly with mathematics completed a math test in the presence of either 2 males (stereotype threat condition) or 2 females (control condition). Participants were also shown a subliminal noise generator before the test. They were told that the machine would either increase arousal and heart rate or that it would not produce any adverse effects. The female participants in the stereotype threat condition performed worse on the math test only when they were told that the subliminal noise generator would not produce any adverse effects. The participants who were told that the machine would increase arousal attributed their difficulties to the noise generator, and did not experience stereotype threat effects. Providing subliminal noise generators are not practical in normal testing situations. However, recent studies have used different techniques to produce the same effect. Johns, Inzlict and Schmader (2008) eliminated stereotype threat effect by telling individuals under stereotype threat conditions that the anxious feelings they were experiencing would not hinder their performance, and might even improve their performance. Attributing difficulties to external factors is a strategy that can be utilized effectively in order to reduce stereotype threat.
Stereotype threat is highly pervasive and is not restricted to people with certain characteristics or backgrounds. Although it can affect anyone, certain individuals are more susceptible to the effects of stereotype threat than others. The effects of stereotype threat make it a serious problem that needs to be addressed. The literature has