|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
EXCEPTION TO THE RULE:
CAN TEST ADMINISTRATORS PREVENT STEREOTYPE THREAT?
RASHIDA WILLIAMS BROWN
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
O 2007 Rashida Williams Brown
To all who nurtured my faith, dreams, and scholarship
The completion of this dissertation and my graduate studies would not have been possible
without the favor of God and support and guidance of a number of truly amazing people.
I owe a great deal of gratitude to the members of my dissertation committee, Nancy
Waldron, Tracy Linderholm, Tina Smith-Bonahue, and Vivian Correa. Their endless wisdom,
insight, and belief in my abilities as a scholar have been instrumental to my intellectual and
interpersonal development as a graduate student. They have all served as role models, exemplars
of female academicians that are committed to fostering future professionals.
I must express my profound appreciation for all the love and support that I have received
from my family and friends throughout the years, the last six especially. From the first day of
graduate school, they have provided the emotional, financial and spiritual support needed to run
this mental marathon. I share this victory with them all. In addition, I sincerely thank my
husband, Barrington, for selflessly investing in my dream. I simply could not have made it this
far without him. I love him infinitely, he is my water.
Finally, this dissertation is in loving memory of my grandmother, Sara Wright, who
constantly prayed for me to "finish writing my book." The prayers of the righteous availeth
TABLE OF CONTENTS
ACKNOWLEDGEMENT S ............ ..... ._ ...............4....
LIST OF TABLES ............ ..... .__ ...............7...
AB S TRAC T ......_ ................. ............_........8
1 INTRODUCTION ................. ...............9.......... ......
Mediating Variables of Stereotype Threat ................. ...............15........... ...
A nxiety .............. ...............15....
W working M emory .............. ...............16....
Evaluation Apprehension .............. ...............18....
Effort ................ .......... ........ ...............19......
Stereotype Threat in Actual Settings ................ ...............20........... ...
Reducing Stereotype Threat .............. ...............22....
Proposed Model .............. ... ...............27...
Study Overview and Hypothesis .............. ...............33....
2 M ETHODS .............. ...............36....
Participants and Design .............. ...............36....
Procedure .............. ...............37....
3 RE SULT S .............. ...............40....
Omnibus Analysis of Test Performance ................. ...............40........... ...
Test Experience Questionnaire ................ ...............40........... ....
Stereotype Threat Questionnaire Items .............. ...............41....
Test Administrator Status Questionnaire Item .............. ...............41....
4 DI SCUS SSION ................. ...............43................
A PARTICIPANT ELIGIBILITY SURVEY ................. ...............50................
B CONSENT FORM ................. ...............51........... ....
C TEST ADMINISTRATOR SCRIPT FOR non-ROLE MODEL CONDITION ................... .53
D TEST ADMINISTRATOR SCRIPT FOR ROLE MODEL CONDITION ...........................54
E STUDY TEST QUESTIONS .............. ...............55....
F TEST EXPERIENCE SURVEY .............. ...............60....
LIST OF REFERENCES .........__.. ..... ._ __ ...............61....
BIOGRAPHICAL SKETCH .............. ...............66....
LIST OF TABLES
2-1 Demographics of sample ........... ......._._ ...............39..
3-1 Analysis of covariance for test performance .............. ...............42....
3-2 Test score (% correct) as a function of student ethnicity, test administrator ethnicity,
and test administrator status ........._._.._......_.. ...............42....
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
EXCEPTION TO THE RULE:
CAN TEST ADMINISTRATORS PREVENT STEREOTYPE THREAT?
Rashida Williams Brown
Chair: Nancy Waldron
Cochair: Tracy Linderholm
Major: School Psychology
Negative stereotypes about intellectual abilities can produce a threat that upsets the
performance of stigmatized groups, such as women in mathematics and students of color in
academic domains, a phenomenon called stereotype threat. There is empirical evidence that
supports stereotype threat and reveals that priming negative stereotypes lead to lower intellectual
performance in stigmatized group members. This study explores factors that can prevent
stereotype threat. Specifically, the study examines whether stereotype threat can be prevented
when an African-American role model administers a test to African-American students. Portions
of a standardized achievement test were administered to European-American and African-
American college students to determine the impact of test administrators on standardized test
performance. Test administrators were found not to impact test performance or reduce
The disparity in school achievement and standardized test performance among racial and
ethnic groups is extensively documented. These gaps in school achievement and standardized
test performance have been sizeable and persistent. The disturbing trend that European-
Americans perform better than African-American and Latino students from primary school
through college is demonstrated across various measures of academic success, such as test
scores, grades, dropout rates, and college matriculation rates (Aronson, 2002). The Scholastic
Aptitude Test (SAT) mean verbal and math scores are 434 and 429 for African-Americans; 454
and 465 for Mexican Americans; 459 and 456 for Puerto Ricans; 458 and 463 for other
Hispanics; 527 and 536 for European-Americans, respectively (Sathy, Barbuti, & Mattern,
2006). Yet, standardized tests are used to make educational decisions that have great
implications for one's life. Standardized tests, such as the SAT and Graduate Record Exam
(GRE) are used for admission decisions and determining scholarship eligibility in higher
education. The No Child Left Behind Act has increased the value and frequency of standard
testing in public schools. Across many states, high school students will be required to pass a
standardized test to graduate. The importance of standardized test scores has led to great
concern about group differences in performance.
The correlation of level of education with income and quality of life has consistently been
demonstrated (Kozol, 1991). Lower test scores on standardized test hinders acceptance to many
postsecondary institutions, reducing the number of high school graduates of color matriculating
into universities. There is substantial evidence that students of color score lower on standardized
tests than European-Americans and females perform lower than males on quantitative portions of
standardized tests (Aronson, Quinn, & Spencer, 1998). Moreover, there is substantial evidence
that students of color score lower on the GRE, which is a test that determines admission to
graduate programs. Thus, scores on the GRE may be a maj or contribution to the significant
underrepresentation of students of color in doctoral programs and professional schools (e.g.,
medical and veterinary). Specifically, the Educational Testing Service (2003) documented the
GRE verbal mean score of 394 and GRE quantitative mean score of 425 for African-Americans,
428 and 483 for Mexican Americans, 408 and 478 for Puerto Ricans, 441 and 499 for other
Hispanics, and 494 and 556 for European-Americans, respectively. Thus, reducing the racial
disparity among standardized test performance will increase access to postsecondary education
for students of color and increase the likelihood of higher income, given that income is correlated
to high salaries (Kozol, 1991) and quality of life.
Researchers have examined various explanations for the discrepancy in standardized test
scores across various ethnic groups that range from environmental factors, such as
socioeconomic disadvantage (McKay, Doverspike, Bowen-Hilton, & McKay, 2003) to
individual factors, such as genetics (Osborne, 2001). The most notorious and heavily disputed
explanation has been based on the premise that some ethnic groups are intellectually inferior by
nature. Herrnstein and Murray (1994) in their controversial book, 7lhe Bell Curve, argued that
there are genetically determined differences in intellectual ability that are demonstrated by the
persistent gap in test scores and school achievement. Many rej ect the notion of group intellectual
inferiority and acknowledge that a combination of sociocultural factors (e.g., poor schooling,
poverty, parental education level, etc.) hinders African-American and Latino students from
optimally achieving in schools (Nieto, 2004). Historically, students of color have been denied
economic and educational resources, subsequently leading to underperformance in certain
academic domains (Kozol, 2001).
Socialization differences among ethnic groups have also been proposed as exacerbating the
achievement gap. Ogbu (1978) argues that certain students of color underperform academically
because there is a mismatch between school and home culture. Certain students of color that
experience continuous, systemic discrimination perceive schooling as providing unequal
educational opportunities and do not endorse certain dominant group behaviors that tend to
facilitate academic achievement, prohibiting them from optimally performing in academic
settings. Specifically, Ogbu purports that African-Americans that are "involuntary minorities,"
American citizens as a result of enslavement, do not subscribe to White middleclass values such
as education (1987). Consequently, the academic failure of African-American students can be
interpreted as reactionary behavior toward the dominant culture because they equate school
success with losing their own cultural identity. Despite various established theories, the
persistent gap in standardized test scores has driven research that further explores factors that
impair the test performance of students of color.
Research suggests that negative stereotypes can impair test performance (Demo & Parker,
1987). Negative stereotypes about members of historically stigmatized groups continue to exist
despite a social and legal climate that explicitly discourages expressions of prejudice and racism
(Wittenbrink, Judd, & Park, 1997). Steele and his colleagues (Steele, 1997; Steele & Aronson,
1995, 1998; Spencer, Steele, & Quinn 1999) have identified the phenomenon of stereotype threat
which purports that cognitive performance is impaired as a result of pressure that one will
confirm a negative stereotype about a group to which one belongs. Specifically, stereotype
threat is defined as "the event of a negative stereotype about a group to which one belongs
becoming self-relevant, usually as a plausible interpretation for something one is doing, for an
experience one is having, or for a situation one is in, that has relevance to one's self-definition"
(Steele, 1997, p.617). It is important to note that the definition emphasizes that a stereotype must
be self-relevant, pertaining to a characteristic that is essential to one's identity, to be threatening.
Stereotype threat is a situational threat that depends heavily on the person's connection with the
stereotype-relevant domain, such as academics. For students in which academic achievement is
a major component of their identity, a negative stereotype regarding intellectual inferiority is
threatening because it pertains to a domain that is personally meaningful.
For stereotype threat to occur a person must strongly identify with the respective group and
the task must be from a domain that the person can identify with. For example, Shih, Pittinsky,
and Ambady (1999) found that performance on a math test depended on the social identity
(Asian or female) that was made salient to students by presenting various ethnic or gender
stereotypes. Aligning with the stereotypes, activating ethnicity improved performance, while
activating gender impaired performance. In a similar study, Hispanic American females
significantly underperformed European-American females and Hispanic American males on a
math test when described as diagnostic of intellectual ability (Gonzales, Blanton, & Williams,
2002). The authors purport that the stereotype threat effects result from the double-minority
status of Hispanic females. Many students encompass a "cultural self" (Markus & Kitayama,
1995), which can serve as a level of interdependence with others of their race and gender. This
interdependence can lead to the perceived responsibility to help their culture thrive. When faced
with a stereotype, even if the person believes it is false, the cultural self can foster an intense
need to dispel the stereotype and activate the awareness that an individual's poor performance
can be generalized to one's group (Cohen & Garcia, 2005). Those who belong to a group with a
negative stereotype about their intellectual performance fear failure and embarrassment, as well
as confirming the negative stereotype. The combination of these factors makes it likely that
academic performance will be negatively affected in threatened groups.
Negative stereotypes that allege the intellectual inferiority of African-American and Latino
students are especially hazardous because stereotypes are well known and intelligence is
universally valued. In schools there is often a climate of stereotype awareness that can provoke
African-American and Latino students to feel suspicious, to fear confirming negative
stereotypes, and to question if they belong in environments of academic domain (Aronson,
2002). Thus, students' reactions to stereotyping can have a significant impact on their academic
performance (Steele, 1997; Steele & Aronson, 1995, 1998; Spencer, Steele, & Quinn 1999).
Steele and Aronson (1995) demonstrated that telling African-American and European-
American undergraduates that a test made up of difficult items from the GRE that were
evaluative of their intellectual ability depressed the test performance of African-American
students. In contrast, when African-Americans and European-Americans were instructed that the
test was non-evaluative the difference in test scores between groups diminished. Even in non-
evaluative testing conditions race was a salient factor. In another study, African-American
students performed significantly lower than European-American student when they were merely
asked to indicate their race on a demographic survey prior to beginning the test. Therefore,
Steele and Aronson (1995) demonstrated that simply requiring a student to indicate their race can
trigger a defense of one' s academic competence by activating a stereotype about that specific
racial group's intellectual abilities.
Stereotype threat can also pertain to stereotypes regarding gender, socioeconomic status,
and age. Stereotype threat has been found in women in male-dominated domains such as math
or science (Spencer et al., 1999). Male and female undergraduates with majors that required
strong math abilities were administered difficult items of the quantitative portion of the GRE.
When participants were instructed that the test was evaluative of their math abilities, female
students performed significantly lower. In contrast, when participants were instructed that the
test was non-evaluative of their math abilities female students performed just as well as their
male counterparts. The effect of stereotype threat on women's math performance has been
replicated in numerous studies (e.g., Inzlicht & Ben-Zeev, 2000; Keller & Dauenheimer, 2003;
O'Brien & Crandall, 2003; Sekaquaptewa & Tompson, 2003).
Similarly, Hess and colleagues (2003) demonstrated stereotype threat effects with elderly
individuals when stereotypes alleging a decline in mental abilities in the elderly were discussed
in the description of the task. Specifically, elderly participants performed worse on a measure of
short-term memory when prompted with stereotypes regarding senility and age than when they
were prompted with stereotypes regarding age and wisdom. Moreover, there is empirical
evidence that stereotypes regarding intellectual inferiority can have an additive effect on
There is also evidence that a person does not have to be a member of a stigmatized
minority group to succumb to the pressures of stereotype threat. European-American male
undergraduates who maj ored in math were administered a difficult math test and in certain
conditions they were instructed that the purpose of the study was to determine why Asian males
perform better on that specific measure. When prompted with the stereotype of Asian
mathematical superiority, participants solved fewer problems and reported lower self-confidence
(Aronson et al., 1999). In addition, European-Americans performed significantly lower than
African-Americans on a motor task when described as measuring natural athletic ability (Stone,
Lynch, Sjomeling, & Darley, 1999). Therefore, anyone who identifies with a group that is
targeted by a stereotype regarding superiority of a comparable group can experience stereotype
Mediating Variables of Stereotype Threat
Numerous studies have demonstrated the effects of stereotype threat. However, the
underlying processes or mediating variables of stereotype threat that impairs performance is still
unclear. Researchers have examined various mediating processes of stereotype threat including
anxiety, reduction in working memory capacity, evaluation apprehension, and effort.
Anxiety is one of the first mechanisms proposed as a mediator of stereotype threat (Steele
& Aronson, 1995). It is hypothesized that anxiety can stem from not only the rigor of the test but
also from the pursuit of disproving the negative stereotype, which could be considered a social
goal. Most students in completing an intellectual activity do not just see the task as an academic
goal, but attach social goals to the task as well (Urdan & Machr, 1995). The pursuit of multiple
goals then increases the student' s anxiety to such a level that it impairs his or her performance.
With anxiety's ability to spring up from several different areas, it is a logical candidate for a
mediator of stereotype threat.
A few studies have demonstrated that stereotype threat conditions produce higher levels of
self-reported anxiety, using the Spielberger (1970) state-trait anxiety instrument (Spencer et al.,
1999; Stone, Lynch, Sjomeling, & Darley, 1999). However, anxiety was found to only have a
marginal effect and could not be deduced as a complete mediator of stereotype threat. Osborne
(2001) demonstrated that anxiety partially mediates stereotype threat effects in African-
American and Latino students by measuring levels of self-reported anxiety in high school
students on three different achievement tests. It should be noted that no mediation effect was
found in Native American students and the group differences in anxiety for African-American
and Latino students were minimally significant. In the same study, it was also found that self-
reported anxiety partially mediated stereotype threat effects on women's math performance
(Osborne, 2001; Spencer et al., 1999).
Anxiety has also been examined as a mediator of stereotype threat by direct, physiological
assessments of anxiety instead of solely relying on self-reports. For example, Blascovich, Steele,
Spencer, and Quinn (2001) found that African-Americans experienced increases in blood
pressure under stereotype threat conditions. Although a few studies have indicated anxiety as a
partial mediator of stereotype threat (e.g., Blascovich et al., 2001; Osborne, 2001; Spencer et al.,
1999), there are numerous studies that have not found evidence of anxiety as a partial or
complete mediator (Gonzales et al., 2002; Keller & Dauenheimer, 2003; Spencer et. al, 1999:
Steele & Aronson, 1995). Thus, the available evidence is inconsistent regarding whether anxiety
is related to depressed performance under stereotype threat conditions (Smith, 2004).
A reduction in working memory capacity is another underlying process that has been
empirically evaluated as a mechanism of stereotype threat. Working memory integrates new
information received from environmental input with stored information from long-term memory.
Attention helps in the identification and selection of information for further processing in
working memory (Baddeley, 1996). It is hypothesized that attentional resources focused on
dispelling negative stereotypes are no longer available, leading to subsequent difficulty encoding
and retrieving information in long-term memory and the impairment test performance.
Schmader and Johns (2003) investigated the relationship between stereotype threat and working-
memory capacity and demonstrated that students of color and females performed lower on
working memory tasks after the priming of negative stereotypes. In a study regarding ethnicity,
European-American and Latino undergraduates completed the operation span task. The
operation span task is a dual processing test in which participants must evaluate whether a
mathematical equation is correct or incorrect while memorizing words for later recall (Turner &
Engle, 1989). In the stereotype condition, the task was described as measuring intellectual
ability. The task was described as a measure of working memory in the control condition.
Latino students recalled significantly fewer words than European-American students under
stereotype threat conditions, indicating that the Latino students' working memory was disrupted
when they perceived that their intellectual ability was being assessed. In a second study
regarding gender, male and female students completed the operation span task. In the stereotype
threat condition, the task was described as measuring mathematic aptitude that is related to group
differences on standardized math tests. Female students recalled significantly fewer words than
male students on the operation span task, demonstrating reduced working-memory capacity.
Thus, the evidence that is currently available on working memory as a mediating factor is
Additionally, there is research that examines the effect of stereotype threat on other forms
of cognitive functioning. For example, Quinn and Spencer (2001) examined whether stereotype
threat disrupts one's ability to formulate the problem-solving strategies that are required to
complete complex mental tasks, such as advanced math problems. Females under stereotype
threat conditions had greater difficulty formulating problem-solving strategies while completing
mathematical word problems than when in reduced threat conditions. Similarly, Croizet and
colleagues (2004) revealed that an increased mental load, which is indicated by reductions in
heart rate variability, can mediate the negative effects of stereotype threat. Heart rate usually
increases during mental tasks and heart rate variability, interval fluctuations between consecutive
heart beats decreases consistently with mental load (Jorna, 1992). Based upon the stereotype that
students maj oring in psychology have lower intelligence than students maj oring in physical
sciences, psychology students were compared with engineering students regarding their
performance on an intelligence test. In the stereotype condition, the test was described as
measuring the intellectual ability required in logical and mathematical reasoning. In the control
condition, the test was described as a non-diagnostic laboratory exercise. Psychology students
performed lower than engineering students under stereotype threat conditions. Moreover,
psychology students experienced a larger increase in heart rate and a larger decrease in heart rate
variability than the engineering students. These results indicate that when presented with a task
that is evaluative, stigmatized group members experienced increased mental load. Evidence
regarding whether depressed problem-solving ability or increased mental load is related to lower
performance under stereotype threat conditions is emerging. However, more research in these
areas is needed to establish substantial evidence.
Evaluation apprehension has also been considered as a mediating factor of stereotype
threat. It is hypothesized that stereotype threat impairs one's performance by generating a
debilitating self-consciousness about how others are evaluating him or her (Smith, 2004). Steele
and Aronson (1995) assessed the evaluation apprehension hypothesis by having African-
American and Latino participants under stereotype threat condition indicate the degree to which
providing their race concerned them and found marginal effects. Evaluation apprehension was
further assessed by requiring European-American males threatened by the stereotype of Asian
intellectual superiority to complete a self-report scale that rated how much they worried about
the experimenter judging them following a difficult math test (Aronson et al., 1999). A majority
of the participants under the threat condition reported worrying about what the experimenter
thought of them. However, the self-report scale used in the study has been challenged as an
inadequate measure of evaluation apprehension (Smith, 2004). Due to marginal results and
questions regarding the methodology of assessment across various studies, evaluation
apprehension is a less viable hypothesis for the mediation of stereotype threat effects.
The withdrawal and overextension of effort has been examined as mediators of the
negative effects of stereotype threat. Steele and Aronson (1995) propose that individuals under
stereotype threat experience self-doubt which reduces motivation, leading to less effort and
subsequent poor performance. In contrast, it has been proposed that individuals under stereotype
threat expend more effort than non-threatened counterparts in the pursuit of dispelling the
stereotype, leading to overexertion and subsequent poor performance (Aronson et al., 1999).
Numerous studies that operationalize effort as the number of items attempted or the time spent
on test items found that effort does not mediate the negative effects of stereotype threat on
academic performance (Aronson et al., 1999, Spencer et al. 1999, Ambady et al., 2001). Self-
reports have also been used to evaluate effort as a mediator (Gonzales et al., 2002; Keller &
Dauenheimer, 2003; Stone et al. 1999). However, effort was found not to mediate stereotype
threat. Since the discovery of stereotype threat effort has been identified as a factor that
intuitively mediates stereotype threat. Therefore, various methodologies are arising in the
examination of effort as a mediator of stereotype threat.
Researchers have examined various mediating variables of stereotype threat.
Unfortunately, no single construct has been reliably identified as a complete mediator of
stereotype threat. The complexity of operationalizing and examining the various hypothesized
mediating variables of stereotype threat has prevented a consensus. Therefore, how stereotype
threat impairs performance is still unclear. The affective mediators of anxiety and evaluation
apprehension have been examined individually as separate factors. However, both capture the
same feeling of worry about confirming negative stereotypes, resulting in the question of
whether anxiety and evaluation apprehension are the same affect being called two different
names. Withdrawal and overextension of effort has also been identified as an affective mediator
of stereotype threat. Yet, difficulty in operationalizing effort has resulted in a lack of consensus
on the appropriate methodological approach to investigating effort as a mediator of stereotype
threat. While research regarding the reduction of working memory capacity is methodologically
sound, working memory capacity does not solely capture the process of stereotype threat.
Stereotype threat is a complex phenomenon that may involve multiple, related affective and
cognitive mediators that impair performance. Specifically, stereotype threat may be comprised
of interrelated affective factors that reduce attention and/or working memory.
Stereotype Threat in Actual Settings
Although empirical evidence that stereotype threat, very few stereotype threat studies have
been conducted in authentic test settings and not in a laboratory (Stricker, 1998; Stricker &
Ward, 1998; Walters, Lee, & Trapani, 2004). Therefore, the question of whether stereotype
threat occurs in authentic testing settings still remains. Stricker (1998) examined whether
requiring high school students to identify their ethnicity and gender prior to taking the exam, a
commonly used subtle method of evoking stereotype threat in laboratory investigations, would
impact performance on the Advanced Placement Calculus AB Examination in an authentic test
administration. Identifying one's ethnicity and gender had no significant effect on the test
performance of African-American or female examinees. Similarly, Stricker and Ward (1998)
examined whether requiring community college students to identify their ethnicity and gender
upon completing the exam would impact performance on the Computerized Placement Tests in
an actual test administration. There were no significant differences between African-American
and female examinees that provided their ethnicity and gender and their counterparts that were
not required to provide that information. Thus, environmental factors that evoke negative effects
of stereotype threat in laboratory settings did not impair performance in an authentic testing
In a similar study, Walters, Lee, and Trapani (2004) examined whether stereotype threat
occurs in authentic test settings without intentional manipulation of the environment to evoke
thoughts of stereotypes. Specifically, they examined whether size, activity level, and social
atmosphere of the test centers, as well as the gender and ethnicity of test administrators,
impacted performance on the GRE. No evidence of stereotype threat was found. Moreover, test
administrator ethnicity and gender had no significantly positive or negative impact on the
performance of African-American or female students. It was hypothesized that one' s typical
environment and exposure to people of similar ethnicity may affect susceptibility to stereotype
threat. In other words, students that are commonly exposed to people of similar ethnicity will
not experience awareness or social comparison of a test administrator of the same ethnicity.
Conversely, students that are rarely exposed to people of similar ethnicity are more likely to be
aware of people of same ethnicity in their environment and perceive a sense of collective
identity. Once more, stereotype threat effects that were found in laboratory settings were not
found in an authentic testing setting.
Administering a standardized exam in any setting can cause evaluation apprehension, yet
laboratory settings are more threatening for examinees. Taking an exam in an authentic testing
situation will result in a score that is used in making educational decisions (e.g., graduation
eligibility, college admission decisions, etc.), while taking an exam in a laboratory setting will
result in a score that is used for research regarding group differences in test performance. Thus,
taking a standardized exam in an authentic testing situation has personally relevant implications,
resulting in less susceptibility of stereotype threat. Conversely, taking a standardized exam in a
laboratory for research purposes has group implications, increasing susceptibility of stereotype
threat for stigmatized group members.
Reducing Stereotype Threat
Despite the continuing debate surrounding the underlying processes of stereotype threat
and its existence in actual test settings, another body of research has emerged that focuses on
identifying interventions for preventing or reducing stereotype threat. Steele (1997) provides the
"wise schooling theory" which purports that various principles of multicultural education can
prevent or reduce stereotype threat. He recommends multicultural education as a means of
addressing the achievement gap because it can build school identification in students that do not
identify with academics and reduce stereotype threat for students that are concerned about their
academic performance. Valuing multiple perspectives in the classroom can demonstrate an
appreciation of diversity and promote an environment in which stereotypes are less likely to be
used and validated (Steele, 2004).
Providing role models can also dispel stereotypes that students may endorse and is a maj or
component of the wise schooling theory. Specifically, exposing students to "people from the
stereotype-threatened group who have been successful in the domain" can demonstrate that
excelling in that domain is possible (Steele, 2004, p. 695). Marx and Roman (2002)
demonstrated that a female role model prevented stereotype threat effects among female
undergraduates. Role model was operationalized as a female that is highly competent in math.
To display competence and foster status as a role model, the test administrator presented the
math test as a difficult exam she created and that she would provide feedback regarding each test
taker' s strength and weaknesses following the exam. Under stereotype threat conditions, female
students performed as well as male students on a difficult math test when the test administrator
was a female role model. In contrast, female students performed lower than their male
counterparts on a difficult math test when the test was administered by a male. Therefore,
providing role models to students of stigmatized groups can prevent stereotype threat.
Building positive teacher-student relationships is another component of the wise schooling
theory and is paramount in multicultural education. Stereotype threat evokes the fear that others
in one's environment will doubt his or her abilities. Building relationships with students that are
based in positive regard for one's potential can reduce feelings of conditional acceptance and
performance-based approval by teachers (Nieto, 2004). Cohen, Steele, & Ross (1997)
demonstrated that critical feedback can be highly motivating for African-American students
when coupled with optimism regarding their potential. Thus, positive teacher-student
relationships can decrease stereotype threat.
Taylor and Antony (2000) explored Steele's wise schooling theory through a qualitative,
ethnographic study, which examined how certain universities have reduced stereotype threat
from negatively impacting graduate students of color. They conducted semi-structured, open-
ended interviews with 12 African-American doctoral students from predominantly European-
American universities. Analysis of the interviews revealed converging evidence for the
importance of building positive teacher-student relationships in the reduction of stereotype
threat. Eleven out of 12 students reported having positive relationships with their advisors that
were supportive and reaffirmed their academic value, reducing their susceptibility to stereotype
In contrast to many factors that contribute to the achievement gap, stereotype threat is
situational. Therefore, various modifications in the environment have been considered for
reducing or eliminating stereotype threat. Based upon the premise that an individual must
perceive a collective identity to a stigmatized group in order to succumb to stereotype threat
(Steele & Aronson, 1995), Ambady and colleagues (2004) proposed that individuation prior to
an exam can reduce susceptibility to stereotype threat. Prior to completing a math test, female
undergraduates under the individuation condition were required to fill out a survey of self-
description questions, as well as identify three negative and positive personal traits. Even after
direct priming of stereotypes regarding mathematical inferiority of women, female students
under individuation conditions performed significantly better than their non-individuated
counterparts. The authors concluded that one's identity is multi-faceted and stimulating various
aspects of identity can make a stereotype that is applicable only to one particular social identity
minor and irrelevant.
The role of self-categorization in the reduction of stereotype threat has also been examined.
In one study, female students were required to describe themselves in an open-ended format
prior to completing a difficult math test. Allowing female students to activate more
individualistic aspects of their identity increased their math performance. In a second study, the
role of self-affirmation in the reduction of stereotype threat was also examined. Based upon
Steele's (1997) assertion that stereotype threat is a threat to one's sense of self-adequacy, Croizet
and colleagues (2001) hypothesized that providing participants of stigmatized groups the
opportunity to affirm positive aspects of their self-concept immediately before taking a test
should reduce the perception of threat, subsequently alleviating negative consequences on
performance. Female students in the experimental condition were required to complete
questionnaires that implied that they were participating in the study because they were good
students, evoking self-affirmations of academic competence. Female students under the standard
stereotype threat condition performed lower than their male counterparts. In contrast, female
students under the experimental stereotype threat condition that completed the self-affirmation
questionnaire performance increased, even higher than their male counterparts. Thus,
emphasizing one's positive self-image through positive self-affirmations may reduce the
negative effects of stereotype threat.
In a similar investigation of stereotype interventions, Aronson, Fried, and Good (2002)
proposed that reconceptualizing intelligence as expandable instead of fixed can reduce stereotype
threat by lessening the power of stereotypes that purport group intellectual inferiority. In the
experimental condition of their study, undergraduates participated in a mock pen pal program in
which they wrote three letters to encourage a seventh-grade student that was experiencing
academic difficulties. Participants were prompted to write letters that encouraged students to
work hard regardless of their environments and to stress that intelligence is malleable,
comparable to a muscle that grows through learning. The purpose of the mock pen pal program
was to persuade the writers themselves to believe the malleability of intelligence. African-
American and European-American students that participated in the experimental condition
earned higher grade point averages than their counterparts in the control conditions.
Similarly, an intervention program which focused on promoting the belief that intelligence
is expandable was effective in increasing academic identification and academic achievement in
secondary school students (Good, Aronson, Inzlicht, 2003). Specifically, college students
mentored seventh-grade students in which they were encouraged to attribute academic
difficulties to the novelty of entering secondary school and to view intelligence as expandable,
which can be improved through education. Females and students of color that participated in the
mentor program performed significantly higher on a standardized test than their counterparts in
the control condition.
Researchers have examined various interventions for preventing or reducing stereotype
threat. Steele's wise schooling theory proposes a framework of multicultural education practices
that could foster a strong self-perception that is less susceptible to stereotype threat. However,
Steele's wise schooling theory provides an overall approach in educating students to prevent
stereotype threat and does not suggest specific interventions to implement in the environment
during the actual testing situation. Similarly, reconceptualizing intelligence has been found to
prevent stereotype threat but it has not been investigated as an intervention that is implemented
in a specific testing situation. While improving the overall schooling of students and developing
intervention programs that change students' perceptions of themselves and their abilities over
time can be effective in preventing stereotype threat, greater effort to identify interventions to
reduce stereotype threat when it is most salient, during the actual testing situation, is needed.
Individuation and self-categorization have emerged as specific interventions implemented in the
testing environment. Yet, further research is needed for strong empirical support of either
intervention. Moreover, individuation and self-categorization are similar in the effort to reduce
the sense of collective identity, resulting in the question of whether they are actually the same
intervention in different forms.
There is qualitative support for Steele's wise schooling theory as a whole and emerging
quantitative support of specific components, such as role models. The use of role models in the
testing situation has been found to reduce stereotype threat among female undergraduates (Marx
& Roman, 2002). Further investigation regarding the impact of role models in the testing
situation among students of color is needed. Additionally, defining a role model and
operationalizing role model status need to be carefully considered in future research to reduce
concerns regarding methodology.
Steele' s wise schooling theory (1997) proposes that role models can reduce stereotype
threat by dispelling negative stereotypes and demonstrating that members of a respective
stigmatized group can excel in a domain. The term role model is often used and can be defined
in many ways. For the purpose of this study, role model was operationalized on the basis of
previous research regarding role models and stereotype threat and social identity theory (Tajfel
& Turner, 1986).
Marx and Roman (2002) demonstrated that female role models reduced stereotype threat
effects on the math performance of female undergraduates. They identified three characteristics
that a role model must encompass to be effective in improving academic performance. The
student must be domain-identified and care about their academic performance (Steele, 1997).
Students must perceive the role model as similar in physical appearance and domain
identification (Lockwood & Kunda, 1997). Finally, the success of the role model must be
viewed as attainable. Overall, the effectiveness of a role model relies heavily on a student' s
ability to make social comparisons with the role model.
A person' s identity is complex, constructed of various layers of individual and group
identity. Group identity activates a collective identity that allows an individual to perceive
similarities with other members of a social group. Tajfel and Turner (1986) acknowledge the
individual and group dynamic of a person's identity in the social identity theory (SIT), which
proposes that there are two components on which identity is built. The first component is
personal identity, the unique physical and personality characteristics of an individual. The second
is social identity, which are the memberships an individual has in various groups. SIT presents
an individual's identity as a point on a continuum ranging from personal identity to social
identity, with one's identity at a specific time represented by a single point on the continuum. An
array of variables affect whether personal identity or social identity will be most salient, and
which of the many group memberships will be most prominent in one's social identity.
Additionally, Tajfel and Turner (1986) argued that everyone strives to maintain a positive social
identity, which frequently occurs through social comparison with other social groups. Individuals
tend to categorize their social world into ingroups and outgroups. An individual is motivated to
improve the position of their ingroup in comparison to outgroups which improves the value of
one's personal identity (Schmader, 2002).
Worchel, luzzini, Coutant, and Ivaldi (2000) present an expanded model of identity,
purporting that there are four, instead of two, components that form identity. One component is
personal identity, which like the SIT model includes an individual's physical and personality
characteristics. A second component, which is labeled group membership, encompasses the
social identity end of the continuum of SIT. Group membership includes the categorization of
the social world into groups and perceptions about membership in these groups. Additionally,
social identity is as much about the groups to which one does not belong (outgroups) as the
group to which one does belong ingroupp). The third component is intragroup identity, which
recognizes that individuals occupy positions within their ingroups. Specifically, intragroup
identity includes the status and role one has within a group and the relationship one has with
ingroup members. The fourth and final dimension, group identity, emphasizes the need of the
group to develop an identity of its own. Group identity includes the group's beliefs, values,
history, and reputation in comparison to other groups. Groups strive to maintain their collective
identity, often pressuring individual members to support and represent this identity. In contrast
to the single continuum of SIT, this expanded model proposes that identity operates on all four
components simultaneously. Each component has its own continuum ranging from high to low
salience, the degree of prominence or awareness of a dimension at a specific time. Thus, an
individual's identity at any single point of time is made up of contributions from each of these
Applying Worchel et al.(2000) to the current study, an African-American student with an
African-American role model may experience the following in their levels of identity when
conducting social comparison with the role model: (1) awareness of personal identity may be
triggered by a similarity in skin color with the test administrator, (2) being African-American
can activate awareness of group membership, (3) awareness of intragroup identity may be
activated by the belief that as a university student you are expected to perform better than other
African-Americans that have not acquired a post-secondary education, and (4) awareness of
group identity can be activated by the motivation to positively represent African-Americans. In
this situation, African-American students may experience greater salience in personal identity
and intragroup identity. Specifically, African-American students will be aware of the role model,
focusing more on their similarities with the role model. There will be less focus on the
difference between the ingroup and outgroup and more focus within the ingroup, reducing the
susceptibility of stereotype threat. The less one perceives differences between their ingroup and
outgroups, the less susceptible they are to stereotype threat. African-American role models
reduce the student' s focus on the comparison of their ingroup's performance with the outgroup's
performance. Additionally, African-American role models represent an exception to the
negative stereotype and promote the plausibility of academic success for students of color.
In contrast, when a European-American administers the test to an African-American
student under stereotype threat conditions it creates an environment that emphasizes the
difference in the student's ingroup (African-Americans) from the outgroup (European-
American). The African-American student' s awareness of personal identity will more likely be
triggered by a difference in skin color from fellow test-takers and the test administrator. In
addition, awareness of group identity will be activated by concern with dispelling the negative
stereotype of intellectual inferiority among African-Americans. Stereotype threat activates the
awareness that an individual's poor performance can be generalized to one's group, impairing
academic performance (Cohen & Garcia, 2005).
A person's awareness of others during testing has been used as a prime for stereotype
threat in previous studies (Aronson, Lustina, Good, & Keough, 1999; Ben-Zeev, Fein, Inzlicht,
2005; Schmader et al., 2003; Spencer et al., 1999; Steele et al., 1995). It is proposed that this
same awareness of others can also prevent or reduce stereotype threat. Occasionally, the
administrator is used as a prime in stereotype threat research, either as a member of a group
regarded as intellectually superior or the person that describes the task as diagnostic of
intellectual ability (Schmader & Johns, 2003; Spencer, Steele, & Quinn, 1999; Steele &
Aronson, 1995). Thus, the administrator can have a significant effect on the test-taker.
Consistent with Steele (1997) and Marx and Roman (2002) that purports providing role
models can dispel stereotypes that students may endorse, the current study proposes that a test
administrator that serves a positive role model can prevent or reduce stereotype threat among
African-Americans. Group identity can allow the test-taker to relate to the administrator as a
person of the same stigmatized group that is successful academically, contradicting the
stereotype. In other words, the role can elicit a sense of confidence and/or self-efficacy that
optimizes the student's performance.
In addition, the ethnicity of the test administrator can influence a student of color' s level of
evaluation apprehension. Walters, Shepperd, and Brown (2003) found that African-American
undergraduates performed worse than European-American undergraduates on a standardized test
when the administrator was European-American, but equal to European-American
undergraduates when the administrator was African-American. They proposed that students of
color underperformed when the administrator was European-American due to evaluation
apprehension. Specifically, students who care about their academic competence are usually
motivated to portray a competent image to the test administrator. Students of color may believe
that a European-American test administrator endorses stereotypes of group intellectual
inferiority, evoking evaluation apprehension, distraction, and subsequent poor performance. In
contrast, students of color may believe that the African-American administrator does not endorse
negative stereotypes and will evaluate each test taker on his or her individual competence.
Converging evidence was provided by Brown and Dobbins (2004) who found that students
of color held higher performance expectations and expected a fair evaluation in class when they
anticipated evaluation by an instructor of the same ethnicity. In contrast, students of color held
lower performance expectations when they anticipated evaluation by a European-American
instructor. Therefore, it is proposed that an African-American test administrator can also reduce
stereotype threat in African-American students through increased performance expectations.
Students of color may assume that because the African-American administrator has been
academically successful he or she does not endorse stereotypes of group intellectual inferiority,
believes that the test takers are academically competent, and will evaluate each test-taker on their
individual performance. In addition, positive role models of the same ethnicity may reduce
stereotype threat in students of color by serving as an exception to the negative stereotype. As an
exception to the stereotype, the role model challenges the plausibility of the negative stereotype
and promotes that stereotypes are generalizations and do not have to apply to every member of
the stigmatized group. Moreover, positive role models of the same ethnicity and/or gender
further represent an attainable goal which motivates students to work harder and achieve
academically (Lockwood & Kunda, 2002; Zirkel, 2002).
The current study attempted to clarify which features of the test administrator are needed to
reduce stereotype threat. Walters et al. (2003) found that a test administrator of similar ethnicity,
perceived through similar appearance, can reduce stereotype threat. Conversely, Marx and
Roman (2002) purported that students must perceive the role model as similar in physical
appearance and domain identification, as well as the success of the role model as attainable for
the role model to effectively reduce stereotype threat. Both studies demonstrate that similar
physical appearance activates collective identity between the test administrator and participants.
Consequently, the question of whether the other features of a role model that are outlined in
Marx and Roman (2002) are needed for a test administrator to serve as a role model and reduce
stereotype threat. Thus, this study examined whether results of Marx and Roman (2002) were a
function of role model status or just similar physical appearance of test administrator.
Further investigation of the role of test administrator in the reduction of stereotype threat
using a sample of African-American undergraduates from a predominantly European-American
(be consistent) university would be beneficial. A maj ority of studies that have demonstrated
negative effects of stereotype threat among students of color have been conducted at
predominantly European-American universities (Aronson, Quinn, & Spencer, 1998; McKay et
al., 2003; Steele, 1997; Steele, 2004; Steele & Aronson, 1995, Steele & Aronson, 1998; Walters,
Shepperd, & Brown, 2003), an environment in which students of color consistently struggle
academically and socially (Allen, 1992; Flowers, 2002). Demonstrating the effects of test
administrator ethnicity on student performance can have great implications in the development of
interventions and support programs for students of color, as well as the recruitment and retention
of faculty of color, at predominantly European-American universities.
Study Overview and Hypothesis
The purpose of this study was to explore factors that can prevent stereotype threat, a
phenomenon in which negative stereotypes about intellectual abilities can produce a threat that
upsets the performance of women in mathematics and students of color in academic domains
(Steele, 1997). The specific research question for this study was as follows:
Is similar domain identification and attainable goal status required for an African-
American test administrator to serve as a role model to prevent stereotype threat among
Steele' s wise schooling theory (1997) proposes that various principles of multicultural education
can prevent or reduce stereotype threat, including providing role models to students of
stigmatized groups. Role models can reduce stereotype threat by dispelling negative stereotypes
and demonstrating that members of a respective stigmatized group can excel in a domain (Steele,
1997). Thus, this study examined whether role models can reduce stereotype threat among
students of color, which has been shown to mediate the effects of stereotype threat among
women in the domain of mathematics (Marx & Roman, 2002).
The effectiveness of a role model relies heavily upon a student' s ability to relate and make
social comparisons with the role model. A student' s social identity is complex, layered with
ethnicity, gender, socioeconomic status, and various other individual and group characteristics
(Worchel et al., 2000). Lockwood and Kunda (1997) purport that students must perceive
similarities in domain identification and attainable goal status in a role model for he or she to be
motivating. Students must also perceive similarity in physical appearance, for it facilitates social
comparisons in ethnicity and gender. Walters, Shepperd, and Brown (2003) found that test
administrators that are similar in ethnicity can prevent stereotype threat in African-American
students. Mere similarity in physical appearance allowed the students to identify the test
administrator as a fellow member of their stigmatized group, positively impacting their academic
performance under stereotype threat conditions. Therefore, this study examined whether similar
domain identification and attainable goal status are required for an African-American test
administrator to serve as a role model and prevent stereotype threat among African-American
students. Specific hypotheses for this study are as follows:
* A 3-way interaction between ethnicity of the administrator, ethnicity of student, and role
model status of administrator is expected. Specifically, ethnic similarity will prevent
* The performance of African-American students, the stigmatized group members, will be
significantly affected by the ethnicity and role model status of the test administrator (Marx
& Roman, 2002; Walters, Shepperd, & Brown, 2003). Specifically, African-American
students are expected to perform to the level of European-American students when the test
administrator is African-American but superior performance is expected when the test
administrator is both African-American and a role model.
* Conversely, European-American students will be unaffected by the ethnicity or role model
status of the test administrator (Marx & Roman, 2002; Walters, Shepperd, & Brown,
These hypotheses were evaluated by manipulating stereotype threat, describing the test as
diagnostic of intellectual ability, among a sample of African-American and European-American
undergraduates using African-American or European-American test administrators and role
models. Participants completed a brief test questionnaire regarding their test experience. Based
on previous laboratory research (Schmader & Johns, 2003; Walters, Shepperd, & Brown, 2003),
the questionnaire included items that assessed whether participants' perception of the test were
consistent with participants of previous laboratory investigations of stereotype threat. In other
words, there were questionnaire items that assessed participants' perception of the test as an
evaluation of group differences. In addition, an item that required participants to estimate the
administrator' s expectation and interpretation of his or her performance was also included to
evaluate the participants' awareness of the test administrator. Furthermore, an item that required
participants to rate their perception of the test administrator as a role model was also included.
Participants and Design
All participants were college students enrolled in a large state university in the south
eastern United States. Participants included 58 African-American and 54 European-American
Educational Psychology students or scholarship students who participated for course credit or
credit towards a volunteer requirement to maintain their scholarships. Additional demographic
information regarding participants is presented in Table 2-1. Participants were recruited for
participation if they reported in a recruitment survey that they were African-American or
European-American and had a neutral or positive attitude toward verbally-oriented classes. The
recruitment survey requested demographic information: race/ethnicity, age, major, SAT score,
and verbal domain identification (Appendix A). Participants had to identify themselves in the
category of European-American and African-American, activating participants' awareness of
group identity. The classification of"Multiracial" was provided in the recruitment survey.
However, only students that identified themselves as solely African-American or European-
American were eligible to participate in the study due to previous research that found that one
who highly identifies with a stigmatized group is most likely to succumb to stereotype threat
(Steele, 1997; Smith & White, 2001).
To assess verbal domain identification participants were asked to respond to the following
statement (Steele, 1995): "Please indicate your typical attitude toward verbally-oriented classes
(literature, language, writing)." Responses were answered on a Likert scale ranging from 1
(strongly dislike) to 5 (strongly like). Similar to previous research (Marx & Roman, 2002;
Walters et al., 2003), only those who responded with a 3 or greater were eligible to participate to
avoid students that dislike verbally-oriented tasks and are unlikely to give maximum effort.
Participants were also asked to provide the number of academic scholarships they have received
to acquire anecdotal information regarding their academic domain identification. Participants
were randomly assigned to one of four conditions in a 2 (test administrator ethnicity: African-
American vs. European-American) x 2 (test administrator status: role model vs. non role model)
x 2 (student ethnicity: African-American vs. European-American) between-subjects factorial
To evoke stereotype threat various actions were taken. Stereotype threat was initially
activated during the recruitment phase in which students were asked to indicate their ethnicity on
the recruitment survey, signifying that ethnicity would be a variable in the study (Appendix A).
During the experimental sessions, the consent form served as a trigger of stereotype threat and
activated group identity, indicating that participants were selected for the study because they
identified themselves as African-American or European-American (Appendix B). Stereotype
threat was also activated through the description of the study as an investigation of verbal
reasoning abilities immediately before the exam (Steele & Aronson, 1995). The ethnic
composition of experimental session groups further triggered stereotype threat. The combination
of African-American and European-American participants in each session activated group
identity and the subsequent social comparisons and awareness of stereotype threat.
An African-American or European-American test administrator conducted each
experimental session with a group of four to seven participants (M = 5.09; SD = 1.23), including
at least two participants of each ethnic group. On arriving for the experiment, participants were
given a consent form and seated in the testing room. At the beginning of the session, the
administrator officially greeted participants and described the study as an investigation of verbal
reasoning abilities (Appendix C). To promote maximum effort in performance participants were
told that they would receive feedback on their verbal strengths and weaknesses at the end of the
session (Steele & Aronson, 1995). Students that are concerned about their academic
performance are more likely to want to know their score and/or how to improve their test
performance. During the role model condition, the test administrator also introduced themselves
as a doctoral candidate during the greeting to create an impression of a role model (Appendix D).
To further foster status as a role model and display competence, the test administrator presented
the test as a difficult exam she created and that she would be provide feedback regarding their
test performance (Marx & Roman, 2002).
Participants completed items from the GRE that are consistent with those administered by
Steele and Aronson (1995; Appendix E). Although participants were offered feedback in the test
description, they did not actually receive feedback on their performance. Upon completing the
test, participants completed a brief questionnaire regarding their test experience (Appendix F).
To test their perception of stereotype threat, participants rated how much they agree with the
following statements from Schmader & Johns (2003): "I am concerned that the researcher will
judge [African-Americans/European-Americans], as a whole, based on my performance on this
test" and "The researcher will think that [African-Americans/European-Americans], as a whole,
have less intellectual ability if I did not do well on this test." To assess the difference of
participants' perception of the test administrator across role model and non-role model
conditions, participants rated how much they agree with the following statement: "I believe the
researcher will be able to provide feedback that will help me on future tests." Responses were
answered on a Likert scale ranging from 1 (strongly disagree) to 7(strongly agree). Participants
were debriefed after completing the questionnaire, receiving the actual purpose of the study and
told that they would not receive feedback on their performance.
Table 2-1. Demographics of sample.
Number of Participants
Total Sample African-American
Omnibus Analysis of Test Performance
Test performance was calculated as a percentage score by dividing the number of problems
answered correctly out of 17 items by 17 and multiplying the remainder by 100. Test
performance scores were analyzed by performing a 2 (student ethnicity: African-American or
European-American) x 2 (test administrator ethnicity: African-American or European-American)
x 2 (test administrator status: role model or non-role model) ANCOVA on test performance,
with self-reported verbal SAT score serving as the covariate. The ANCOVA results are
presented in Table 3-1. This analysis revealed a significant effect for the covariate, F(1, 93) =
25.63, p < .001, partial 82 = .22; no main effect for student ethnicity, F(1, 93) = 1.76, p = .19,
partial 82 = .02; no main effect for test administrator ethnicity, F(1, 93) = .52, p = .47, partial 82 =
.01; and no main effect for test administrator status, F(1, 93) = .01, p = .91, partial 82 = .00. In
addition, no interactions were found, Fs < 1.16, ps > .29. Descriptive statistics for test
performance are presented in Table 3-2.
Test Experience Questionnaire
As a manipulation check to verify perception of stereotype threat and the difference in
perception of the test administrator across role model and non-role model conditions, participants
completed a brief questionnaire regarding their test experience. The questionnaire was
comprised of two questions regarding their perception of stereotype threat and one question
assessing perception of the administrator. Participants responded using a Likert scale ranging
from 1 (strongly disagree) to 7 (strongly agree). Each questionnaire item was analyzed using a 2
(student ethnicity: African-American or European-American) x 2 (test administrator ethnicity:
African-American or European-American) x 2 (test administrator status: role model or non-role
Stereotype Threat Questionnaire Items
Analysis of the extent to which participants were concerned that the test administrator
would judge their respective ethnic group, as a whole, based on their performance yielded a main
effect for student ethnicity, F(1, 104) = 22.17, p < .01. As expected, European-American
students (M~= 2.06, SD = 1.24) were less concerned than African-American students (M~= 3.48,
SD = 1.82) that the researcher would judge their respective ethnic group based on their test
performance. No main effect for test administrator ethnicity, F(1, 104) = .08, p = .78, partial 82 =
.00, and no main effect for test administrator status, F(1, 104) = .64, p = .43, partial 82 = .01,
were found. In addition, no interactions were found, Fs < 1.84, ps > .18.
Analysis of the extent to which participants were concerned that the test administrator
would perceive their respective ethnic group as having less intellectual ability if he or she did not
do well on the test revealed a main effect for student ethnicity, F(1, 104) = 16.74, p < .01.
European-American students (M~= 1.89, SD = 1.16) were less concerned than African-American
students (M~= 3.10, SD = 1.75) that the researcher would perceive their respective ethnic group
as having less intellectual ability if they performed poorly. No main effect for test administrator
ethnicity, F(1, 104) = .07, p = .80, partial 82 = .00, and no main effect for test administrator
status, F(1, 104) = .53, p = .47, partial 82 = .01, were found. In addition, no interactions were
found, Fs < 2. 18, ps > .14.
Test Administrator Status Questionnaire Item
Analysis of the extent to which participants believed the test administrator would be able
to provide useful feedback yielded a main effect for test administrator status, F(1, 104) = 11.43,
p < .01. Test administrators in the role model condition (M~= 5.41, SD = .93) were viewed as
more likely to provide useful feedback than test proctors in the non-role model condition (M~=
4.81, SD = .98). Thus, it is argued that role model status was effectively operationalized and
presented in the role model condition. In addition, a main effect for student ethnicity, F(1, 104)
= 5.21, p = .024, was also found. African-American students(M~= 5.33, SD = 1.0) were more
likely than European-American students (M~= 4.91, SD = .96) to anticipate receiving helpful
feedback from the test administrator. No main effect for test administrator ethnicity was found,
F(1, 104) = .00, p = .99, partial 82 = .00. In addition, no interactions were found, Fs < 1.98, ps >
Table 3-1. Analysis of covariance for test performance
Source Df F E2
Verbal SAT (covariate) 1 25.63 .22 .00
Student ethnicity (SE) 1 1.76 .02 .19
Test administrator ethnicity (TE) 1 .52 .01 .47
Test administrator status (TS) 1 .01 .00 .91
SE x TE 1 .08 .00 .78
SE x TS 1 .01 .00 .91
TE x TS 1 1.16 .01 .29
SE x TE x TS 1 .70 .01 .40
Error 93 (165.38)
Note. Value enclosed in parentheses represents mean square error.
Table 3-2. Test score (% correct) as a function of student ethnicity, test administrator ethnicity,
and test administrator status
Test administrator status
Role model Non-role model
AA EA AA EA
admini strator admini strator admini strator admini strator
AA student 45.27 (14.89) 42.21 (13.02) 42.47 (11.21) 41.27 (18.04)
EA student 40.85 (11.89) 38.45 (17.48 36.46 (13.72) 36.00 (15.94)
Note. AA = African-American, EA = European-American Values enclosed mn parentheses represent standard deviations
The purpose of this study was to explore factors that can prevent stereotype threat, a
phenomenon in which negative stereotypes about intellectual abilities can produce a threat that
upsets the performance of women in mathematics and students of color in academic domains
(Steele, 1997). The primary intent of this study was to examine whether similar domain
identification and attainable goal status are required for an African-American test administrator
to serve as a role model to prevent stereotype threat among African-American students. The
current study extends previous research regarding stereotype threat by attempting to clarify
which features of the test administrator are needed to reduce stereotype threat, just similar
ethnicity (Walters et al., 2003) or similar ethnicity, domain identification, and attainable status of
the test administrator conj ointly (Marx & Roman, 2002).
African-American participants were expected to perform to the level of European-
American participants when the test administrator was African-American, demonstrating the
inhibition of stereotype threat. Additionally, African-American participants were expected to
perform better than European-American participants when the test administrator was an African-
American role model. However, no evidence of stereotype threat or its reduction was found with
respect to test performance. Similar ethnicity or role model status of the test administrator did
not lead to an increase in test performance for African-American participants.
These findings were inconsistent with expected results that were based on previous
stereotype threat research in laboratory settings (Marx & Roman, 2002; Walters et al, 2003).
Conversely, these findings were consistent with previous stereotype threat research conducted in
actual test settings that found no evidence of stereotype threat (Stricker, 1998; Stricker & Ward,
1998; Walters, Lee, & Trapani, 2004). While there was evidence that African-Americans were
more concerned than European-Americans that the test administrator would perceive their
respective ethnic group as having less intellectual ability if he or she did not do well on the test,
the concern did not lead to a disruption in their performance. Although African-Americans were
more likely than European-Americans to perceive the role model as knowledgeable and credible,
their awareness of the role model did not lead to significantly better performance. Although
nonsignifieant differences in test performance were not unexpected, these Eindings are
remarkable because they demonstrate that some African-American undergraduates do not
experience impaired test performance even when they experience evaluation apprehension.
While impairment of test performance is an integral component of stereotype threat, the current
Endings suggest that some African-American students experience the fear of confirming
negative stereotypes, but not to the extent that it impairs test performance. Thus, future studies
should continue to address under what circumstances groups who often report experiencing
stereotype threat are resilient to its effects during testing. Further investigation of test
environment and interpersonal factors (i.e. prior educational experiences, self-efficacy, etc.) as
forms of stereotype threat prevention is warranted.
It can be speculated that there was no evidence of stereotype threat in the current study
because the environment was not like previous laboratory research. Experimental sessions were
held in a conference room, with four to seven participants sitting at a conference table to
complete the exam. The conference room was a less academic and intimidating environment
than a classroom or computer lab where previous laboratory research was conducted (Marx &
Roman, 2002; Walters et al., 2003), possibly reducing the susceptibility to stereotype threat.
There are additional explanations for the nonsignificant findings. Similar to the previous
study in an authentic testing situation by Stricker (1998) that did not find evidence of stereotype
threat, participants were required to provide their ethnicity upon completing the exam instead of
prior to taking the exam. Although students were required to identify their ethnicity in the
recruitment survey and the consent form reminded participants that they were selected for the
study because they identified themselves as African-American or European-American,
participants may not have perceived ethnicity to be their most salient attribute during the exam.
Not having to provide any personal information in the exam packet prior to the exam may have
created a sense of anonymity, possibly reducing the fear of evaluation and susceptibility of
stereotype threat. In addition, the exam was described as an examination of personal factors
involved in solving problems that require verbal reasoning abilities, which may have shifted
participants' awareness away from their group identity to their personal identity during the exam.
Interestingly enough, African-American participants were more concerned than European-
American participants about their performance being perceived as representative of their
respective group. However, participants provided those responses on the questionnaire
following the exam in which they were required to provide their ethnicity, which may have
shifted participants' awareness to their group identity. Requiring participants to provide their
ethnicity explicitly activates group identity and stereotype awareness while the description of the
exam is more subtle. The difference in the intensity of the environmental cues may be a possible
explanation for why there was some evidence of stereotype threat on the questionnaire but not on
test performance. Previous studies (Marx & Roman, 2002; Walters et al., 2003) that did not
require participants to provide their ethnicity before completing the exam and found evidence of
stereotype threat described the exam as evaluative of intellectual abilities or mathematical
ability, explicitly evoking awareness of negative stereotypes regarding group intellectual
In contrast to previous laboratory studies that involved freshmen and sophomores from
introductory psychology courses (Steele, 1995; Steele & Aronson, 1995; Walters et al., 2003;
Marx & Roman, 2002), a maj ority of the sample was comprised of upper-lever students across
various college majors. Thus, the sample was similar to stereotype threat research conducted at
actual GRE testing sites in which upper level students were taking the GRE for admission to
graduate school. Upper-level students may be less susceptible to stereotype threat because they
have had years of collegiate experience that affirms their self-concept as intelligent and
academically competent. A portion of the sample was comprised of African-American and
European-American students with academic scholarships. Scholarship students may be less
susceptible to stereotype threat because they have earned scholarships for their superior test
performance, affirming confidence in their test-taking abilities. Students with positive self-
concepts regarding test performance may be less likely to fear performing poorly and confirming
negative stereotypes regarding their respective groups. Moreover, the African-American
participants were undergraduates of a predominantly White university who have experienced
success in academic settings in which there were few African-Americans. Thus, the African-
American participants may be desensitized to academic experiences in which they are in the
minority, subsequently reducing susceptibility to stereotype threat.
The current study intended to examine whether similar domain identification and attainable
goal status is required for a test administrator to prevent stereotype threat. However, neither test
administrator ethnicity nor role model status were found to significantly impact test performance.
Thus, the nonsignificant differences in test performance across conditions did not lead to
clarification of whether test administrator ethnicity exclusively prevents stereotype threat
(Walters et al., 2003) or whether attainable goal status and similar domain identification also
needed to prevent stereotype threat (Marx & Roman, 2002)
Prior to discussing implications of this study, it is important to mention its limitations. The
most significant limitation was the diversity and size of the sample. Although the sample was
comprised of a balanced number of African-American and European-American participants, the
maj ority of the sample was upper-level students and a notable portion of the sample was
scholarship students, all of whom were comfortable or expressed a preference for verbal tasks.
To increase the generalizabilty of the study, additional lower-level and non-scholarship students
could be included in the sample. In addition, a greater number of female students participated in
the study, possibly skewing results. To be eligible to participate in the study, participants had to
indicate a neutral or positive attitude towards verbally-oriented courses. Requiring such
preference possibly reduced the number of male participants that are more likely to prefer math-
Some stereotype threat studies regarding differences among ethnic groups primarily
focused on accumulating a balanced sample of African-American and European-American
participants, resulting in minimal focus on a balance of gender across participants and a sample
of more females than males (McKay et al., 2003; Steele & Aronson, 1995; Walters et al., 2003;
Walters et al., 2004). Future research could include the systematic investigation of test
administrator and student gender and ethnicity as factors in the prevention of stereotype threat
among African-American students. Specifically, greater effort to balance gender and ethnicity of
the sample could lead to a better understanding of stereotype threat and increase generalizability.
Although there were an adequate number of participants, increasing the gender college level
diversity of participants would make the sample more representative of the university population.
Furthermore, data analysis of test performance involved self-reported verbal SAT scores as
a covariate. Self-reported test scores have to be interpreted with caution due to the fact that they
may not be as valid and reliable as actual test scores. It has been found in previous research that
students who did not perform well on standardized tests were more likely to report false test
scores that are higher than their actual score (Kuncel, Crede, & Thomas, 2005). It is important to
note that there were no significant ethnic group differences in the self-reported verbal SAT
This study addresses various issues that are relevant to the disparity in school achievement
and standardized test performance among racial and ethnic groups. This study did not reveal
evidence of stereotype threat that was demonstrated in previous research as a possible
explanation for such disparity. These Eindings may be a function of the limitations in this study
or an indicator that test administrator ethnicity or role model status do not induce or reduce
stereotype threat. These Eindings align with emerging research that has demonstrated a lack of
evidence for stereotype threat (Stricker, 1998; Stricker & Ward, 1998; Walters et al., 2004). It is
possible that stereotype threat only occurs in laboratory settings and testing environments that
evoke awareness of ethnic differences among examinees. Future research should investigate
possible explanations for the differences in laboratory and Hield study results. Further research is
needed to determine which environmental factors evoke and inhibit stereotype threat.
The lack of evidence of stereotype threat in this study demonstrates that some African-
American students are not susceptible to stereotype threat. While previous research has
examined specific interventions or modifications in the testing in environment as a means to
prevent or reduce stereotype threat (Ambady et al., 2004, Aronson et al., 2002; Croizet et al.,
2001; Marx & Roman, 2002; Steele, 1997; Steele, 2004; Walters et al. 2003), further research is
needed to identify what interpersonal factors, besides ethnicity or gender, make students less
susceptible to stereotype threat. Identifying personal characteristics that decrease susceptibility
to stereotype threat can enhance the understanding of the phenomenon.
There is still a need for more knowledge of factors contributing to the persistent
achievement gap in our society. To reduce and eventually end the disparity in school
achievement and standardized test performance among racial and ethnic groups, we first must
identify factors that lead to this gap. In view of the potential impact of stereotype threat findings
on the disparity in standardized test performance and concern for equity in standardized testing,
it is crucial that there is more investigation of the role stereotype threat plays in the test
performance of students of color and methods that might ameliorate any adverse affects that
stereotype threat may have on test performance.
PARTICIPANT ELIGIBILITY SURVEY
Please circle the race/ethnicity that you most identify with.
White/Europ ean-Ameri can
Black (non-Hi spani c)/Afri can-Ameri can
If you took the SAT, please circle your approximate verbal score.
Did not take SAT
Please indicate the number of academic scholarships you have ever received.
Please indicate your maj or.
Please indicate your typical attitude toward verbally-oriented classes (literature, language,
Strongly Dislike Neutral Strongly Like
1 2 3 4 5
You are invited to participate in a research study that investigates personal factors involved
in performance on problems requiring reading and verbal reasoning abilities. You chose to
participate in the study as fulfilling part of your course or scholarship requirements. You were
selected as a possible participant because you agreed to participate on a voluntary basis and have
expressed a neutral or positive attitude towards verbally-oriented courses. We ask that you read
this form and ask any questions you may have before agreeing to be in the study. Rashida
Brown, doctoral student of the Educational Psychology department, is conducting the study. Dr.
Tracy Linderholm, Assistant Professor of Educational Psychology, University of Florida, and Dr.
Nancy Waldron, Associate Professor of School Psychology, are supervising this study.
The purpose of this study is to understand personal factors that impact performance on
problems that require reading and verbal reasoning abilities. The results of this study will have
educational implications in terms of providing guidance for standardized test performance.
You will be asked to complete a 17-item test of verbal problems. Following the test you
will be asked to complete a short survey regarding your test experience. The entire procedure
will require less than an hour of your time.
Risks and Benefits of Participating in the Study:
There are no direct benefits to participation in this study. The only possible risk of
participating in this study is mental fatigue. The probability of this risk' s occurrence is extremely
Scholars: Participation in research earns course or scholarship credit. Participants can earn
volunteer credits toward maintaining their scholarships as dictated by the Office of Admissions.
Education course students: Participants can earn extra credits determined by the individual
course instructor. Instructors will be notified that you have participated.
Voluntary Nature of the Study:
Your decision whether or not to participate will not affect your current or future relations
with the University of Florida. If you decide to participate, you are free to withdraw at any time
without affecting those relationships and without forfeiting the compensation (course or
scholarship credit) that was agreed upon.
The records of this study will be kept private, and your identity will be kept confidential to
the extent required by law. Your information will be assigned a code number. The list connecting
your name to this number will be kept in a locked file and can be accessed only by the
researcher. When the study is completed and the data have been analyzed, the list will be
destroyed. In addition, the experimental results will be analyzed in terms of averages across
participants rather than in terms of individual performances. In any form of report we might
publish, we will not include any information that will make it possible to identify a research
Contacts and Questions:
The researcher conducting this study is Rashida Brown, a doctoral student. Dr. Tracy
Linderholm, Assistant Professor of Educational Psychology, University of Florida, and Dr.
Nancy Waldron, Associate Professor of School Psychology, are supervising this study. You may
ask any questions you have now. If you have questions later, please contact the experimenter,
Rashida Brown, Educational Psychology, (3 52) 246-5 100 or Dr. Tracy Linderholm, Educational
Psychology, (352) 392-0723 ext. 241 or Dr. Nancy Waldron, Educational Psychology, (352)
392-0723 ext. 232.
If you would like to speak to someone other than the researcher about your concerns
regarding the study and/or your rights as a research participant, please contact UFIRB office,
P.O. Box 112250, University of Florida, Gainesville, FL 32611-2250; phone (352) 392-0433.
You will be given a copy of this form to keep for your records.
Statement of Consent:
I have read the above information and have obtained a copy of this form. I have asked
questions and have received satisfactory answers. I consent to participate in the study.
Signature of Participant Date
Signature of Investigator Date
TEST ADMINISTRATOR SCRIPT FOR NON-ROLE MODEL CONDITION
My name is Thank you for your participation. The purpose of this study is
to investigate various personal factors involved in performance on problems requiring reading
and verbal reasoning abilities. You are asked to complete a 17-item test of verbal problems in a
format identical to the SAT exam and a short survey regarding your test experience. The test and
survey will take approximately 30 minutes of your time. The test is very difficult and you should
expect not to get many of the questions correct. The exam is a genuine test of your verbal
abilities and limitations so that we might better understand the factors involved in both. At the
end of the session, you will be given feedback on your performance that may be helpful to you
for future tests by familiarizing you with some of your strengths and weaknesses in verbal
problem solving. Although you can expect not to get many of the questions correct, please give
your maximum effort for it will help us in our analysis of your verbal ability. Please complete
each question. Since this is an analysis of your verbal ability, we ask that you indicate an answer
in which you guessed by placing a check next to that item number. Once again, it is important
that you give maximum effort. Thank you again for your participation. Are there any questions?
TEST ADMINISTRATOR SCRIPT FOR ROLE MODEL CONDITION
My name is I am a doctoral candidate in the Department of Educational
Psychology at UF and I am researching language processing in young adults for my doctoral
dissertation. Like you, I was once an undergraduate student at UF so I want to thank you for
giving of your time to participate in my dissertation study. The purpose of this study is to
investigate various personal factors involved in performance on problems requiring reading and
verbal reasoning abilities. You are asked to complete a 17-item test of verbal problems that I
have created. The test is in a format identical to the SAT exam. Following the test, you are asked
to complete a short survey regarding your test experience. The test and survey will take
approximately 30 minutes of your time. The test is very difficult and you should expect not to get
many of the questions correct. The exam is a genuine test of your verbal abilities and limitations
so that I might better understand the factors involved in both. At the end of the session, I will
give you feedback on your performance that may be helpful to you for future tests by
familiarizing you with some of your strengths and weaknesses in verbal problem solving.
Although you can expect not to get many of the questions correct, please give your maximum
effort for it will help me in the analysis of your verbal ability. Please complete each question.
Since this is an analysis of your verbal ability, I ask that you indicate an answer in which you
guessed by placing a check next to that item number. Once again, it is important that you give
maximum effort. Thank you again for your participation. Are there any questions?
STUDY TEST QUESTIONS
Directions: Each sentence below has one or two blanks, each blank indicating that something
has been omitted. Beneath the sentence are Hyve lettered words or sets of words. Choose the
word or set of words for each blank that best fits the meaning of the sentence as a whole.
1. The new biological psychiatry does not deny the contributing roles of psychological factors
in mental illness, but that these factors may act as a catalyst on existing physiological conditions
and such illnesses.
2. During periods of social and cultural stability, many art academies are so firmly controlled
by that all real creative work must be done by the
A. dogmatists disenfranchised
B. managers reactionaries
C. reformers dissatisfied
D. impostors academicians
E. specialists elite
3. The first World War began in a context of jargon and verbal delicacy and continued in a
cloud of as as language and literature, skillfully used, could make it.
A. circumlocution literal
B. cliche lucid
C. euphemism impenetrable
D. particularity deliberate
E. subjectivity enthralling
4. While Parker is very outspoken on issues she cares about, she is not ; she concedes
the _of opposing arguments when they expose weaknesses in her own.
A. fickle validity
B. arrogant restraint
C. fanatical strength
D. congenial incompatibility
E. unyielding speciousness
5. Usually the first to spot data that were inconsistent with other findings, in this particular
experiment she let a number of results slip by.
6. Cezanne's delicate watercolor sketches often served as of a subj ect, a way of
gathering fuller knowledge before the artist' s final engagement of the subj ect in an oil painting.
A. an abstraction
B. an enchantment
C. a synthesis
D. a reconnaissance
E. a transcription
7. Liberty is not easy, but far better to be an fox, hungry and threatened on its hill, than a
canary, safe and secure in its cage.
A. unfriendly fragile
B. aging young
C. angry content
D. imperious lethargic
E. unfettered well-fed
8. Johnson never to ignore the standards of decent conduct mandated by company policy
if compliance with instructions from his superiors enabled him to do so, whatever the
effects on his subordinates.
A. deigned tacit
B. attempted half-hearted
C. intended direct
D. scrupled literal
E. wished feigned
Directions: In each of the following questions, a related pair of words or phrases is followed by
five lettered pairs of words or phrases. Select the lettered pair that best expresses a relationship
similar to that expressed in the original pair.
9. didactic : instruct : :
A. pedantic : contend
B. comic :amuse
C. theatrical : applaud
D. imperative : obey
E. rhetorical : recite
10. garrulous : talkative : :
A. suspicious : unreliable
B. cantankerous : obtuse
C. cloying : sweet
D. reflective : insightful
E. prudent : indecisive
11. digressive : statement :
A. connotative : definition
B. slanderous : slur
C. tangential : presupposition
D. biased : opinion
E. circuitous : route
12. chicanery : clever : :
A. expertise : knowledgeable
B. certainty : doubtful
C. gullibility : skeptical
D. machination : heedless
E. tactlessness : truthful
13. pluck : quit : :
A. verve :flinch
B. gall :skimp
C. pride : grovel
D. charm :smile
E. poise :waver
Directions: The passage below is followed by questions based on its content. After reading the
passage, choose the best answer to each question. Answer all questions following the passage on
the basis of what is stated or implied in that passage.
The historian Frederick J. Turner wrote in the 1890s that the agrarian discontent that had
been developing steadily in the United States since about 1870 had been precipitated by the
closing of the internal frontier--that is, the depletion of available new land needed for further
expansion of the American farming system. Not only was Turner's thesis influential at the time,
it was later adapted and elaborated by other scholars, such as John D. Hicks in The Populist
Revolt (1931). Actually, however, new lands were taken up for farming in the United States
throughout and beyond the nineteenth century. In the 1890's, when agrarian discontent had
become most acute, 1,100,000 new farms were settled, which was 500,000 more than had been
settled during the previous decade. After 1890, under the terms of the Homestead Act and its
successors, more new land was taken up for farming than had been taken up for this purpose in
the United States up until that time. It is true that a high proportion of the newly farmed land
was suitable only for grazing and dry farming, but agricultural practices had become sufficiently
advanced to make it possible to increase the profitability of farming by utilizing even these
relatively barren lands.
The emphasis given by both scholars and statesmen to the presumed disappearance of the
American frontier helped to obscure the great importance of changes in the conditions and
consequences of international trade that occurred during the second half of the nineteenth
century. In 1869 the Suez Canal was opened and the first transcontinental railroad in the United
Sates was completed. An extensive network of telegraph and telephone communications was
spun: Europe was connected by submarine cable with the United States in 1886 and with South
America in 1874. By about 1870 improvements in agricultural technology made possible the full
exploitation of areas that were most suitable for extensive farming on a mechanized basis. Huge
tracts of land were being settled and farmed in Argentina, Australia, Canada, and in the
American West, and these areas were j oined with one another and with the countries of Europe
into an interdependent market system. As a consequence, agrarian depressions no longer were
local or national in scope, and they struck several nations whose internal frontiers had not
vanished or were not about to vanish. Between the early 1870s and 1890s, the mounting agrarian
discontent in America paralleled the almost uninterrupted decline in the prices of American
agricultural products on foreign markets. Those staple-growing farmers in the United States who
exhibited the greatest discontent were those who had become most dependent on foreign markets
for sale of their products. Insofar as Americans had been deterred from taking up new land for
farming, it was because market conditions had made this period a perilous time in which to do
14. According to the author, changes in the conditions of international trade resulted in an:
A. underestimation of the amount of new land that was being farmed in the United States
B. underutilization of relatively small but rich plots of land
C. overexpansion of the world transportation network for shipping agricultural products
D. extension of agrarian depressions beyond national boundaries
E. emphasis on the importance of market forces in determining the process of agricultural
15. The author implies that the change in that state of the American farmers morale during the
latter part of the nineteenth century was traceable to the American farmer' s increasing perception
A. costs of cultivating the land were prohibitive within the United States
B. development of the first transcontinental railroad in the United States occurred at the
expense of the American farmer
C. American farming system was about to run out of the new farmland that was required for
D. prices of American agricultural products were deteriorating especially rapidly on
E. proceeds from the sales of American agricultural products on foreign markets were
16. According to the passage, which of the following occurred prior to 1890?
A. Frederick J. Turner' s thesis regarding the American frontier became influential
B. The Homestead Act led to an increase in the amount of newly farmed land in the United
C. The manufacturers of technologically advanced agricultural machinery rapidly increased
their marketing efforts
D. Direct lines of communication were constructed between the United States and South
E. Technological advances made it fruitful to farm extensively on a mechanized basis
17. The author's argument implies that, compared to the yearly price changes that actually
occurred on foreign agricultural markets during the 1880s, American farmers would have most
preferred yearly price changes that were
A. much smaller and in the same direction
B. much smaller but in the opposite direction
C. slightly smaller and in the same direction
D. similar in size but in the opposite direction
E. slightly greater and in the same direction
TEST EXPERIENCE SURVEY
Please rate how much you agree with the following statements.
I am concerned that the researcher will judge [African-Americans/European-Americans], as a
whole, based on my performance on this test.
Strongly Disagree Neutral Strongly Agree
1 2 3 4 5 6 7
The researcher will think that [African-Americans/European-Americans], as a whole, have less
intellectual ability if I did not do well on this test.
Strongly Disagree Neutral Strongly Agree
1 2 3 4 5 6 7
I believe the researcher will be able to provide feedback that will help me on future tests.
Strongly Disagree Neutral Strongly Agree
1 2 3 4 5 6 7
LIST OF REFERENCES
Allen, W. R. (1992). The color of success: African American college student outcomes at
predominantly white and historically black public college and universities. H~arvard
Educational Review, 64 (1, 26-44.
Ambady, N., Paik, S. K., Steele, J., Owen-Smith, A., Mitchell, J. P. (2004). Deflecting negative
self-relevant stereotype activation: The effects of individuation. Journal ofExperintental
Psychology, 40, 401-408.
Ambady, N., Shih, M., Kim, A., and Pittinsky, T. L. (2001). Stereotype susceptibility in
children: Effects of identity activation on quantitative performance. Psychological
Science, 12, 385-390.
Aronson, J. (2002). Stereotype threat: Contending and coping with unnerving expectations. In J.
Aronson (Ed.), Insproving academic achievement: Inspact of psychological factors on
education (pp. 279-301). New York, NY: Academic Press.
Aronson, J., Fried, C. B., and 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.
Aronson, J., Lustina, M. J., Good, C., Keough, K., Steele, C. M., & Brown, J. (1999).
When white men can't do math: Necessary and sufficient factors in stereotype threat.
Journal ofExperintental Social Psychology, 35, 29-46.
Aronson, J., Quinn, D. M., & Spencer, S. J. (1998). Stereotype threat and the academic
underperformance of minorities and women. In J. K. Swim & C. Stangor (Eds.),
Prejudice: The target 's perspective (83-103).. New York, NY: Academic Press.
Baddeley, A. (1996). Exploring the central executive. The Quarterly Journal ofExperintental
Psychology, 49A (1, 5-28.
Ben-Zeev, T., Fein, S., & Inzlicht, M. (2005). Arousal and stereotype threat. Journal of
Experimental Social Psychology, 41, 174-18 1.
Blascovich, J., Spencer, S. J., Quinn, D., & Steele, C. M. (2001). African Americans and high
blood pressure: The role of stereotype threat. Psychological Science, 12, 225-229.
Brown, L. M., & Dobbins, H. (2004). Students' of color and European American students'
stigma-relevant perceptions of university instructors. Journal of SociallIssues, 60(1I),
Cohen, G. L., & Garcia, J. (2005). "I am us": Negative stereotypes as collective threats. Journal
ofPersonality and Social Psychology, 89(4), 566-582.
Croizet, J., Despres, G., Gauzins, M., Huguet, P., Leyens, J., & Meot, A. (2004). Stereotype
threat undermines intellectual performance by triggering a disruptive mental load.
Personality and Social Psychology Bulletin, 30(6), 721-73 1.
Croizet, J., Desert, M., Dutrevis, M., & Leyens, J. (2001). Stereotype threat, social
class, gender, and academic under-achievement: when our reputation catches up to us and
takes over. Social Psychology ofEducation, 4, 295-3 10.
Demo, D. H., & Parker, K. D. (1987). Academic achievement and self-esteem among
African-American and White college students. Journal ofSocialPsychology, 127, 345-
Educational Testing Service. (2003). Sex, race, ethnicity, and performance on the GRE R
General Test 2003-2004. Retrieved March 1, 2005, from http:.//ftp.ets.org/pub/
Flowers, L. (2002). The impact of college racial composition on African American students'
academic and social gains: Additional evidence. Journal of College Student
Development, 43(3), 403-410.
Good, C., Aronson, J., & Inzlicht, M. (2003). Improving adolescents' standardized test
performance: An intervention to reduce the effects of stereotype threat. Applied
Developmental Psychology, 24, 645-662.
Gonzales, P. M., Blanton, H., & Williams, K. J. (2002). The effects of stereotype threat and
double-minority status on the test performance of Latino women. Personality and Social
Psychology Bulletin, 28, 659-670.
Herrnstein, R. J., & Murray, C. (1994). The bell curve. Intelligence and class structure in
American hife. New York, NY: Free Press.
Hess, T. M., Auman, C., Colcombe, S. J., & Rahhal, T. A. (2003). The impact of stereotype
threat on age differences in memory performance. Journal of Gerontology, 58, 3-11.
Inzlicht, M., & Ben-Zeev, T. (2000). A threatening intellectual environment: Why women are
susceptible to experience problem-solving deficits in the presence of men. Psychological
Science, 11, 365-371.
Jorna, P. (1992). Spectral analysis of heart rate and psychological state: A review of its validity
as workload index. Biological Psychology, 34, 237-257.
Keller, J., & Dauenheimer, D. (2003). Stereotype threat in the classroom: Dej section mediates the
disrupting threat effecton women's math performance. Personality and Social Psychology
Bulletin, 29, 371-381.
Kozol, J. (1991). Savage inequalities: Children in America 's schools. New York, NY: Harper
Kuncel, N. R., Crede, M. & Thomas, L. L. (2005). The validity of self-reported grade point
averages, class ranks, and test scores: A meta-analysis and review of the literature.
Review of~ducational Resea~rch, 75, 63-82.
Lockwood, P., & Kunda, Z. (2002). Motivation by positive or negative role models: Regulatory
focus determines who will best inspire us. Journal ofPersonality and Social Psychology,
Markus, H. R., & Kitayama, S. (1994). The cultural construction of self and emotion:
implications for social behavior. In S. Kitayama & H. R. Markus (Eds.). Emotion and
culture: Empirical studies of mutual influence (pp. 89-130). Washington, DC: American
Marx, D. M., & Roman, J. S. (2002). Female role models: Protecting women's math test
performance. Personality and Social Psychology Bulletin, 28(9), 1 183-1 193.
McKay, P. F., Doverspike, D., Bowen-Hilton, D., McKay, Q. D. (2003). The effects of
demographic variables and stereotype threat on black/white differences in cognitive
ability test performance. Journal of Business and Psychology, 18, 1-14.
Nieto, S. (2004). Affirming diversity: The sociopolitical context of multicultural education
(4 edition). New York, NY: Longman Publishing Group.
O'Brien, L. T., & Crandall, C. S. (2003). Stereotype threat and arousal: Effects on women's math
performance. Personality and Social Psychology Bulletin, 29, 782-789.
Ogbu, J. U. (1978). Minority education and caste: The American system in cross-cultural
perspective. San Diego: Academic Press.
Ogbu, J. U. (1987). Variability in minority school performance: A problem in search of an
explanation. Anthropology and Education Quarterly, 18(4), 312-334.
Osborne, J. W. (2001). Testing stereotype threat: Does anxiety explain race and sex
differences in achievement? Contemporary Educational Psychology, 26, 291-3 10.
Quinn, D. M., & Spencer, S. J. (2001). The interference of stereotype threat with
women' s generation of mathematical problem-solving strategies. Journal of Social
Issues, 57, 55-71.
Sathy, V., Barbuti, S., Mattern, K. (2006). The new SA T O and trends in test performance.
Retrieved October 14, 2006, from College Board Web site: http: www. collegeboard.
Schmader, T. (2002). Gender identification moderates stereotype threat effects on women's math
performance. Journal ofExperimental Social Psychology, 38, 194-201.
Schmader, T., & Johns, M. (2003). Converging evidence that stereotype threat reduces
working memory capacity. Journal ofPersonality and Social Psychology, 85, 440-452.
Sekaquaptewa, D., & Tompson, M. (2003). Solo status, stereotype threat and performance
expectancies: Their effects on women's performance. Journal ofExperintental Social
Psychology, 39, 68-74.
Shih, M., Pittinsky, T. L., & Ambady, N. (1999). Stereotype susceptibility: Shifts in quantitative
performance from sociocultural identification. Psychological Science, 10, 81-84
Smith, J. L. (2004). Understanding the process of stereotype threat: A review of
mediational variables and new performance goal directions. Educational Psychology
Review, 16 (3), 177-206).
Smith, C. E., & Hopkins, R. (2004). Mitigating the impact of stereotypes on academic
performance: The effects of cultural identity and attributions for success among African
American college students. The Western Journal ofBlack Studies, (28)1, 312-321.
Smith, J. L. & White, P. H. (2001). Development of the domain identification measure: a tool
for investigating stereotype threat effects. Educational andPsychological 2ea~surentent,
Spencer, S. J., Steele, C. M., & Quinn, D. M. (1999). Stereotype threat and women's
math performance. Journal ofExperintental Social Psychology, 35, 4-28.
Spielberger, C. D., Gorsuch, R. L., and Lushene, R. E. (1970). Manual for the State-Trait
Anxiety Inventory. Palo Alto, CA: Consulting Psychologist Press.
Steele, C. M. (2004). A threat in the air: How stereotypes shape intellectual identity
and performance. In J. Banks & C. Banks (Eds.), Handbook of research on multicultural
education (2nd edition)(pp. 682-699). San Francisco, CA: Jossey-Bass.
Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity and
performance. American Psychologist, 52, 613-629.
Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of
African Americans. Journal ofPersonality and Social Psychology, 69, 797-8 11.
Steele, C. M., & Aronson, J. (1998). Stereotype threat and the test performance of
academically successful African Americans. In C. Jencks & M. Phillips (Eds.), The
black-white test score gap. (pp. 401-427). Washington, D.C.: Brookings Institution
Stone, J., Lynch, C. I., Sjomeling, M., & Darley, J. M. (1999). Stereotype threat effects on black
and white athletic performance. Journal ofPersonality and Social Psychology, 77, 1213-
Stricker, L. J. (1998). hIquiring about exa~ninees ethnicity and sex: E~ffects on AP Calcuhus AB
Examination performance (College Board Rep. No. 98-01; ETS RR 98-05). New York,
NY: College Board.
Stricker, L. J., & Ward, W. C. (1998). Inquiring about exa~ninees' ethnicity and sex: Effects on
computerized placement tests performance (College Board Rep. No 98-02; ETS RR 98-
09). Princeton, NJ: Educational Testing Service.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In S.
Worchel & W. G. Austin (Eds.). The psychology of inter group relations (pp. 7-24).
Chicago: Nelson Hall.
Taylor, E., & Antony, J. S. (2000). Stereotype threat reduction and wise schooling: Towards the
successful socialization of African American doctoral students in education. The Journal
of Negro Education, 69 (3), 184-198.
Turner, M. L., & Engle, R. W. (1989). Is working memory task dependent? Journal of~entory
and Language, 28, 127-154.
Urdan, T. C., & Machr, M. L. (1995). Beyond a two-goal theory of motivation and achievement:
A case for social goals. Review of~ducationalResea~rch, 65, 213-243.
Walters, A. M., Lee, S., & Trapani, C. (2004). Stereotype Threat, the Test-Center Environment,
and Performance on the GRE General Test (College Board Rep. No 01-03R; ETS RR-
04-37). Princeton, NJ: Educational Testing Service.
Walters, A., Shepperd, J. A., Brown, L. M. (2003). The effect of test administrator ethnicity on
test performance. Unpublished manuscript, University of Florida.
Wittenbrink, B., Judd, C. M., & Park, B. (1997). Evidence for racial prejudice at the implicit
level and its relationship with questionnaire measures. Journal ofPersonality and Social
Psychology, 72, 262-274.
Worchel, S., & Coutant D. (2001). It takes two to tango: Relating group identity to individual
identity within the framework of group development. In M. A. Hogg & R. S. Tindale
(Eds.), Blaciaell handbook of social psychology: Group processes (pp. 461-481).
Maiden, MA: Oxford Blackwell Publishing.
Zirkel, S. (2002.) Is there a place for me? Role models and academic identity among white
students and students of color. Teachers College Record, 104(2), 357-376.
Rashida Williams Brown was born on December 19, 1979 in Miami, Florida. The
youngest of three children, she grew up mostly in Miami, Florida, graduating from Miami
Norland Senior High School in 1997. She earned her B.S. in psychology and her M. Ed. in
school psychology from the University of Florida in 2001 and 2005, respectively. She earned a
Ph.D. in School Psychology in 2007.