<%BANNER%>

Exception to the Rule: Can Test Administrators Prevent Stereotype Threat?


PAGE 1

EXCEPTION TO THE RULE: CAN TEST ADMINISTRATORS PR EVENT STEREOTYPE THREAT? By RASHIDA WILLIAMS BROWN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Rashida Williams Brown

PAGE 3

3 To all who nurtured my faith, dreams, and scholarship

PAGE 4

4 ACKNOWLEDGEMENTS The completion of this dissertat ion and my graduate studies would not have been possible without the favor of God and support and guida nce of a number of truly amazing people. I owe a great deal of gratit ude 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. Th ey have all served as role models, exemplars of female academicians that are committed to fostering future professionals. I must express my profound appreciation for a ll 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 emotiona l, 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 in vesting 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 gra ndmother, Sara Wright, who constantly prayed for me to finish writing my book. The prayers of the righteous availeth much.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGEMENTS.............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 ABSTRACT....................................................................................................................... ..............8 CHAPTER 1 INTRODUCTION................................................................................................................... .9 Mediating Variables of Stereotype Threat..............................................................................15 Anxiety........................................................................................................................ ....15 Working Memory............................................................................................................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 METHODS........................................................................................................................ .....36 Participants and Design........................................................................................................ ..36 Procedure...................................................................................................................... ..........37 3 RESULTS........................................................................................................................ .......40 Omnibus Analysis of Test Performance.................................................................................40 Test Experience Questionnaire...............................................................................................40 Stereotype Threat Questionnaire Items...........................................................................41 Test Administrator Stat us Questionnaire Item................................................................41 4 DISCUSSION..................................................................................................................... ....43 APPENDIX 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

PAGE 6

6 F TEST EXPERIENCE SURVEY............................................................................................60 LIST OF REFERENCES............................................................................................................. ..61 BIOGRAPHICAL SKETCH.........................................................................................................66

PAGE 7

7 LIST OF TABLES Table page 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

PAGE 8

8 Abstract of Dissertation Pres ented 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 PR EVENT STEREOTYPE THREAT? By Rashida Williams Brown May 2007 Chair: Nancy Waldron Cochair: Tracy Linderholm Major: School Psychology Negative stereotypes about intellectual abili ties 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 primi ng negative stereotypes lead to lower intellectual performance in stigmatized group members. Th is study explores factors that can prevent stereotype threat. Specificall y, the study examines whether stereo type threat can be prevented when an African-American role model administers a test to African-American students. Portions of a standardized achievement test were ad ministered to European-American and AfricanAmerican college students to determine the impact of test administrators on standardized test performance. Test administrators were f ound not to impact test performance or reduce stereotype threat.

PAGE 9

9 CHAPTER 1 INTRODUCTION The disparity in school achievement and sta ndardized test performance among racial and ethnic groups is extensively documented. These ga ps in school achievement and standardized test performance have been sizeable and pers istent. The disturbing trend that EuropeanAmericans perform better than African-Ameri can and Latino students from primary school through college is demonstrated across various m easures of academic success, such as test scores, grades, dropout rates, and college matr iculation 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-American s, respectively (Sathy, Barbuti, & Mattern, 2006). Yet, standardized tests are used to make educational decisions that have great implications for ones life. Standardized test s, such as the SAT and Graduate Record Exam (GRE) are used for admission decisions and dete rmining 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 inco me and quality of life has consistently been demonstrated (Kozol, 1991). Lower test scores on standardized test hi nders acceptance to many postsecondary institutions, reducin g the number of high school graduates of color matriculating into universities. There is substantial evidence th at students of color score lower on standardized tests than European-Americans and females perform lower than males on quantitative portions of standardized tests (Ar onson, Quinn, & Spencer, 1998). Moreove r, there is subs tantial evidence

PAGE 10

10 that students of color score lower on the GRE, which is a test that determines admission to graduate programs. Thus, scores on the GRE ma y be a major contribution to the significant underrepresentation of students of color in doct oral programs and pr ofessional schools (e.g., medical and veterinary). Specifically, the Edu cational Testing Service (2003) documented the GRE verbal mean score of 394 and GRE quantita tive 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-American s, respectively. Thus, reducing the racial disparity among standardized test performance will increase access to po stsecondary education for students of color and increase the likelihood of highe r 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 envi ronmental factors, such as socioeconomic disadvantage (McKay, Dove rspike, Bowen-Hilton, & McKay, 2003) to individual factors, such as ge netics (Osborne, 2001). The most notorious and heavily disputed explanation has been based on the premise that some ethnic groups are in tellectually inferior by nature. Herrnstein and Murray ( 1994) in their controversial book, The Bell Curve argued that there are genetically determined differences in intellectual ability that are demonstrated by the persistent gap in test scores and school achieve ment. Many reject the no tion of group intellectual inferiority and acknowledge that a combination of sociocultura l factors (e.g., poor schooling, poverty, parental education leve l, etc.) hinders African-Ameri can and Latino students from optimally achieving in schools (Nie to, 2004). Historically, students of color have been denied economic and educational resources, subsequen tly leading to underperformance in certain academic domains (Kozol, 2001).

PAGE 11

11 Socialization differences among ethnic groups ha ve 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 ho me 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 th em from optimally performing in academic settings. Specifically, Ogbu purports that Afri can-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 beha vior toward the dominant cultu re because they equate school success with losing their own cultural identity. Despite various established theories, the persistent gap in standardized test scores has dr iven research that furthe r explores factors that impair the test performan ce of students of color. Research suggests that negative stereotypes can impair test performance (Demo & Parker, 1987). Negative stereotypes about members of historically stigma tized 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 iden tified 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 belo ngs. Specifically, stereotype threat is defined as the event of a negativ e stereotype about a group to which one belongs becoming self-relevant, usually as a plausible in terpretation for something one is doing, for an experience one is having, or for a situation one is in, that has relevance to ones self-definition

PAGE 12

12 (Steele, 1997, p.617). It is important to note that the definition emphasizes that a stereotype must be self-relevant, pertaining to a ch aracteristic that is essential to ones identity, to be threatening. Stereotype threat is a situationa l threat that depends heavily on the persons connection with the stereotype-relevant domain, such as academics. For students in which academic achievement is a major component of their ident ity, a negative stereotype regard ing intellectual inferiority is threatening because it pertains to a do main that is personally meaningful. For stereotype threat to occu r a person must strongl y 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 va rious ethnic or gender stereotypes. Aligning with the stereotypes, activating ethnicit y 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 inte llectual ability (Gonzal es, Blanton, & Williams, 2002). The authors purport that the stereotype thr eat effects result from the double-minority status of Hispanic females. Many students en compass a cultural self (Markus & Kitayama, 1995), which can serve as a level of interdependen ce with others of their race and gender. This interdependence can lead to the perceived responsibi lity to help their culture thrive. When faced with a stereotype, even if the pe rson believes it is false, the cultural self can foster an intense need to dispel the stereotype and activate the awareness that an individuals poor performance can be generalized to ones group (Cohen & Ga rcia, 2005). Those who belong to a group with a negative stereotype about their intellectual perf ormance fear failure and embarrassment, as well

PAGE 13

13 as confirming the negative stereotype. The co mbination of these factors makes it likely that academic performance will be negatively affected in threatened groups. Negative stereotypes that allege the intellect ual inferiority of Afri can-American and Latino students are especially hazardous because st ereotypes are well known and intelligence is universally valued. In schools th ere is often a climate of stereo type 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 & Ar onson, 1995, 1998; Spencer, Steele, & Quinn 1999). Steele and Aronson (1995) demonstrated th at telling African-American and EuropeanAmerican undergraduates that a test made up of difficult items from the GRE that were evaluative of their intellectual ability depre ssed the test performan ce of African-American students. In contrast, when Af rican-Americans and European-Americans were instructed that the test was non-evaluative the diffe rence in test scores between groups diminished. Even in nonevaluative testing conditions race was a salien t factor. In another study, African-American students performed significantly lower than Europ ean-American student when they were merely asked to indicate their race on a demographic survey prior to be ginning the test. Therefore, Steele and Aronson (1995) demonstrat ed that simply requiring a stude nt to indicate their race can trigger a defense of ones academic competence by activating a stereotype about that specific racial groups intellectual abilities. Stereotype threat can also pe rtain 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 fe male undergraduates with majors that required

PAGE 14

14 strong math abilities were administered difficult items of the quantitativ e 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 womens 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 th e elderly were discussed in the description of the task. Specifically, elderly participants performed worse on a measure of short-term memory when prompted with stereot ypes regarding senility a nd age than when they were prompted with stereotypes regarding ag e and wisdom. Moreover, there is empirical evidence that stereotypes regarding intellectual inferiority can have an additive effect on academic performance. 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 majored in math were admini stered a difficult math test and in certain conditions they were instructed that the purpos e 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 addi tion, European-Americans performe d 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

PAGE 15

15 targeted by a stereotype regard ing superiority of a comparable group can experience stereotype threat. Mediating Variables of Stereotype Threat Numerous studies have demonstrated the eff ects of stereotype threat. However, the underlying processes or mediating va riables of stereotype threat th at impairs performance is still unclear. Researchers have examined various medi ating processes of stereotype threat including anxiety, reduction in working memory capac ity, evaluation apprehension, and effort. Anxiety 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 dispr oving the negative stereotype, whic h 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 we ll (Urdan & Maehr, 1995). The pursuit of multiple goals then increases the students anxiety to such a level that it impairs his or her performance. With anxietys 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 stereot ype threat conditions pr oduce higher levels of self-reported anxiety, using the Sp ielberger (1970) statetrait anxiety instrume nt (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 co mplete mediator of ster eotype threat. Osborne (2001) demonstrated that anxiety partially medi ates stereotype threat effects in AfricanAmerican and Latino students by measuring leve ls of self-reported anxiety in high school students on three different achieveme nt tests. It should be note d that no mediation effect was found in Native American students and the group differences in anxiety for African-American

PAGE 16

16 and Latino students were minimally significant. In the same study, it was also found that selfreported anxiety partially mediat ed stereotype threat effects on womens 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. Al though 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 & Dauenheim er, 2003; Spencer et. al, 1999: Steele & Aronson, 1995). Thus, the available eviden ce is inconsistent regarding whether anxiety is related to depressed performance under stereotype threat conditions (Smith, 2004). Working Memory A reduction in working memory capacity is another underlying process that has been empirically evaluated as a mechanism of stereo type threat. Working memory integrates new information received from envir onmental input with stored inform ation from long-term memory. Attention helps in the identification and selec tion of information for further processing in working memory (Baddeley, 1996). It is hypothes ized that attentiona l resources focused on dispelling negative stereotypes are no longer available, leading to subsequent difficulty encoding and retrieving information in long-term memo ry and the impairment test performance. Schmader and Johns (2003) inves tigated the relationship between stereotype threat and workingmemory capacity and demonstrated that studen ts of color and females performed lower on working memory tasks after the priming of negative stereotypes. In a st udy regarding ethnicity, European-American and Latino undergraduates completed the operation span task. The

PAGE 17

17 operation span task is a dual pr ocessing test in which participants must evaluate whether a mathematical equation is correct or incorrect wh ile 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 signi ficantly fewer words than Eu ropean-American students under stereotype threat conditions, i ndicating 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 comple ted the operation span task. In the stereotype threat condition, the task was desc ribed as measuring mathematic ap titude 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 generally supportive. Additionally, there is research that examines th e effect of stereotype threat on other forms of cognitive functioning. For example, Quinn a nd Spencer (2001) examined whether stereotype threat disrupts ones ability to formulate the problem-solving stra tegies that are required to complete complex mental tasks, such as advanced math problems. Females under stereotype threat conditions had greater difficulty formula ting problem-solving strate gies while completing mathematical word problems than when in redu ced threat conditions. Similarly, Croizet and colleagues (2004) revealed that an increased mental load, whic h is indicated by reductions in heart rate variability, can mediate the negative eff ects of stereotype threat Heart rate usually increases during mental tasks and heart rate variab ility, interval fluctuations between consecutive heart beats decreases cons istently with mental load (Jorna, 1992). Based upon the stereotype that

PAGE 18

18 students majoring in psychology have lower inte lligence than students majoring in physical sciences, psychology students were compared with engineering students regarding their performance on an intelligence test. In the st ereotype condition, the test was described as measuring the intellectual ability required in l ogical and mathematical reasoning. In the control condition, the test was describe d as a non-diagnostic laboratory exercise. Psychology students performed lower than engineering students und er stereotype threat conditions. Moreover, psychology students experienced a larg er increase in heart rate and a larger decrease in heart rate variability than the engineering students. These re sults indicate that when presented with a task that is evaluative, stigmatized group members e xperienced increased ment al load. Evidence regarding whether depressed problem-solving ability or increased mental load is related to lower performance under stereotype thr eat conditions is emerging. Howe ver, more research in these areas is needed to establish substantial evidence. Evaluation Apprehension Evaluation apprehension has also been consid ered as a mediating f actor 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 ap prehension hypothesis by having AfricanAmerican and Latino participants under stereotype thre at condition indicate the degree to which providing their race concerned them and found marg inal 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 (A ronson et al., 1999). A majority of the participants under the th reat condition reported worrying about what the experimenter thought of them. However, the self-report scal e used in the study has been challenged as an

PAGE 19

19 inadequate measure of evaluation apprehension (Smith, 2004). Due to marginal results and questions regarding the methodology of asse ssment across various studies, evaluation apprehension is a less viable hypothesis for th e mediation of stereot ype threat effects. Effort 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 se lf-doubt which reduces motivati on, leading to less effort and subsequent poor performance. In contrast, it has been proposed th at individuals under stereotype threat expend more effort than non-threatened counterparts in the pursuit of dispelling the stereotype, leading to overex ertion and subsequent poor perf ormance (Aronson et al., 1999). Numerous studies that operationa lize effort as the number of ite ms 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). Selfreports have also been used to evaluate effo rt as a mediator (Gonzal es et al., 2002; Keller & Dauenheimer, 2003; Stone et al. 1999). However, effort was found not to mediate stereotype threat. Since the discovery of st ereotype threat effort has been identified as a factor that intuitively mediates stereotype threat. Ther efore, various methodologies are arising in the examination of effort as a medi ator of stereotype threat. Researchers have examined various media ting variables of stereotype threat. Unfortunately, no single construc t has been reliably identified as a complete mediator of stereotype threat. The comple xity of operationalizing and exam ining the various hypothesized mediating variables of stereotype threat has prev ented 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

PAGE 20

20 same feeling of worry about confirming negativ e stereotypes, resulti ng 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 operationa lizing effort has resulted in a lack of consensus on the appropriate methodological appr oach to investigating effort as a mediator of stereotype threat. While research regarding the reducti on of working memory capacity is methodologically sound, working memory capacity does not solely cap ture 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 th reat may be comprised of interrelated affective f actors 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 (Str icker, 1998; Stricker & Ward, 1998; Walters, Lee, & Trapani, 2004). Ther efore, the question of whether stereotype threat occurs in authentic tes ting settings still remains. St ricker (1998) examined whether requiring high school students to identify their et hnicity 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 Calc ulus AB Examination in an authentic test administration. Identifying ones ethnicity a nd gender had no significant effect on the test performance of African-American or female ex aminees. Similarly, Stricker and Ward (1998) examined whether requiring community college st udents to identify their ethnicity and gender upon completing the exam would impact performa nce on the Computerized Placement Tests in an actual test administration. There were no si gnificant differences between African-American and female examinees that provided their ethnici ty and gender and their counterparts that were

PAGE 21

21 not required to provide that information. Thus, environmental factors that evoke negative effects of stereotype threat in laborat ory settings did not impair perfor mance in an authentic testing situation. In a similar study, Walters, Lee, and Trapani (2 004) examined whether stereotype threat occurs in authentic test settings without inte ntional manipulation of the environment to evoke thoughts of stereotypes. Specifi cally, they examined whether si ze, activity level, and social atmosphere of the test centers, as well as th e 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 signi ficantly positive or negative impact on the performance of African-American or female stud ents. It was hypothesize d that ones typical environment and exposure to people of similar ethni city may affect susceptibility to stereotype threat. In other words, students that are comm only 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 et hnicity are more likely to be aware of people of same ethnic ity in their environment and pe rceive a sense of collective identity. Once more, stereotype th reat 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 thre atening for examinees. Taking an exam in an authentic testing situation will result in a scor e that is used in making edu cational 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 rega rding group differences in test performance. Thus, taking a standardized exam in an authentic testi ng situation has personally relevant implications,

PAGE 22

22 resulting in less susceptibility of stereotype threat. Conversely, taking a standardized exam in a laboratory for research purposes has group implicat ions, increasing susceptib ility 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 resear ch has emerged that focuses on identifying interventions for preventing or reduci ng stereotype threat. St eele (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 thr eat for students that are concerned about their academic performance. Valuing multiple perspe ctives in the classroom can demonstrate an appreciation of diversity and pr omote an environment in which stereotypes are less likely to be used and validated (Steele, 2004). Providing role models can also dispel stereotypes that student s may endorse and is a major component of the wise schooling theory. Speci fically, exposing student s to people from the stereotype-threatened group who ha ve been successful in the domain can demonstrate that excelling in that domain is possible (St eele, 2004, p. 695). Marx and Roman (2002) demonstrated that a female role model prev ented 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 ro le model, the test administrator presented the math test as a difficult exam she created and that she would provide feedback regarding each test takers strength and weaknesses following the ex am. Under stereotype th reat conditions, female students performed as well as male students on a di fficult math test when the test administrator

PAGE 23

23 was a female role model. In contrast, female students performed lower than their male counterparts on a difficult math test when the te st was administered by a male. Therefore, providing role models to students of stigma tized groups can preven t 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 ones environment will doubt his or her abilities Building relationships with students that are based in positive regard for ones potential can reduce feelings of conditional acceptance and performance-based approval by teachers (N ieto, 2004). Cohen, Steele, & Ross (1997) demonstrated that critical feedback can be highly motivating for African-American students when coupled with optimism regarding thei r potential. Thus, positive teacher-student relationships can decrease stereotype threat. Taylor and Antony (2000) explor ed Steeles wise schooling th eory through a qualitative, ethnographic study, which examined how certain uni versities have reduced stereotype threat from negatively impacting graduate students of color. They conducted semi-structured, openended interviews with 12 African-American doc toral students from predominantly EuropeanAmerican universities. Analysis of the inte rviews revealed converging evidence for the importance of building positive teacher-student re lationships 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 threat. In contrast to many factors that contribute to the achievement gap, st ereotype threat is situational. Therefore, various modifications in the environment have been considered for reducing or eliminating stereot ype threat. Based upon the premis e that an individual must

PAGE 24

24 perceive a collective identity to a stigmatized gr oup in order to succumb to stereotype threat (Steele & Aronson, 1995), Ambady and colleagues ( 2004) proposed that in dividuation prior to an exam can reduce susceptibility to stereotype th reat. Prior to completing a math test, female undergraduates under the individua tion condition were required to fill out a survey of selfdescription questions, as well as identify three ne gative and positive personal traits. Even after direct priming of stereotypes regarding mathematical inferior ity of women, female students under individuation conditions performed signifi cantly better than their non-individuated counterparts. The authors concluded that ones 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 th reat has also been examined. In one study, female students we re required to describe themse lves 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 perf ormance. In a second study, the role of self-affirmation in the reduction of stereoty pe threat was also examined. Based upon Steeles (1997) assertion that stereotype threat is a threat to ones sense of self-adequacy, Croizet and colleagues (2001) hypothesized that provid ing participants of stigmatized groups the opportunity to affirm positive as pects of their self-concept imme diately before taking a test should reduce the perception of threat, subse quently alleviating negative consequences on performance. Female students in the experi mental condition were required to complete questionnaires that implied that they were pa rticipating in the study because they were good students, evoking self-affirmations of academic co mpetence. Female students under the standard stereotype threat conditi on performed lower than their male c ounterparts. In contrast, female

PAGE 25

25 students under the experimental stereotype threat condition that completed the self-affirmation questionnaire performance increased, even hi gher than their male counterparts. Thus, emphasizing ones 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 in telligence as expandable instead of fixed can reduce stereotype threat by lessening the power of stereotypes that purport group in tellectual 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 a nd 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 belie ve the malleability of intelligence. AfricanAmerican and European-American students that participated in the experimental condition earned higher grade point averages than th eir counterpart s in the control conditions. Similarly, an intervention program which focu sed on promoting the belief that intelligence is expandable was effective in increasing academic identification and academic achievement in secondary school students (Good, Aronson, Inzlic ht, 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 a nd students of color that participated in the mentor program performed significan tly higher on a standardized te st than their counterparts in the control condition.

PAGE 26

26 Researchers have examined various interven tions for preventing or reducing stereotype threat. Steeles 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, Steeles wise schooling theory provides an overa ll approach in educating students to prevent stereotype threat and do es 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 inve stigated as an intervention that is implemented in a specific testing situation. While improving the overall schoo ling of students and developing intervention programs that change students perc eptions 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-categorizat ion have emerged as specific interventions implemented in the testing environment. Yet, furt her research is needed for stro ng empirical support of either intervention. Moreover, individu ation and self-categorization are similar in the effort to reduce the sense of collective identity, resulting in th e question of whether they are actually the same intervention in different forms. There is qualitative support for Steeles wi se 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 stereo type threat among female undergraduates (Marx & Roman, 2002). Further investigat ion regarding the impact of role models in the testing situation among students of color is neede d. Additionally, defini ng a role model and operationalizing role model status need to be care fully considered in future research to reduce concerns regarding methodology.

PAGE 27

27 Proposed Model Steeles wise schooling theory (1997) proposes that role models can reduce stereotype threat by dispelling negative stereotypes and demonstrating that me mbers 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, ro le 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 id entified three characteristics that a role model must encompass to be eff ective 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 ro le model must be viewed as attainable. Overall, the effectiven ess of a role model relie s heavily on a students ability to make social comparisons with the role model. A persons identity is complex, constructed of various layers of individual and group identity. Group identity activates a collective identity that allo ws an individual to perceive similarities with other members of a social group. Tajfel and Turner (1986) acknowledge the individual and group dynamic of a persons identity in the social identity theory (SIT), which proposes that there are two components on whic h 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 vari ous groups. SIT presents an individuals identity as a point on a conti nuum ranging from personal identity to social identity, with ones identity at a specific time represented by a single point on the continuum. An

PAGE 28

28 array of variables affect whether personal identity or social identity will be most salient, and which of the many group memberships will be mo st prominent in ones social identity. Additionally, Tajfel and Turner ( 1986) argued that ever yone 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 compar ison to outgroups which improves the value of ones personal identity (Schmader, 2002). Worchel, Iuzzini, Coutant, and Ivaldi ( 2000) present an expanded model of identity, purporting that there are four, instead of two, co mponents that form identity. One component is personal identity, which like the SIT model includ es an individuals phys ical 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 th ese 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 (ingroup). The thir d component is intrag roup identity, which recognizes that individuals occupy positions wi thin their ingroups. Specifically, intragroup identity includes the status and role one has within a group and the re lationship one has with ingroup members. The fourth and final dimensi on, group identity, emphasiz es the need of the group to develop an identity of its own. Group identity includes the groups beliefs, values, history, and reputation in comparis on to other groups. Groups strive to maintain their collective identity, often pressuring individu al members to support and represen t this identity. In contrast to the single continuum of SIT, this expanded m odel proposes that identity operates on all four components simultaneously. Each component has its own continuum ranging from high to low

PAGE 29

29 salience, the degree of prominence or awareness of a dimension at a specific time. Thus, an individuals identity at any singl e point of time is made up of contributions from each of these dimensions. Applying Worchel et al.(2000) to the current study, an Africa n-American student with an African-American role model may experience the fo llowing in their levels of identity when conducting social comparison with the role model: (1) awareness of person al identity may be triggered by a similarity in skin color with the test administra tor, (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 expect ed to perform better than other African-Americans that have not acquired a pos t-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 expe rience 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 ro le 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 thre at. The less one perceives diffe rences between their ingroup and outgroups, the less susceptible they are to stereo type threat. African-A merican role models reduce the students focus on the comparison of their ingroups performa nce with the outgroups performance. Additionally, African-American role models represent an exception to the negative stereotype and promote the plausibil ity of academic success for students of color. In contrast, when a European-American admi nisters the test to an African-American student under stereotype thre at conditions it creates an e nvironment that emphasizes the difference in the students ingroup (Afri can-Americans) from the outgroup (European-

PAGE 30

30 American). The African-American students awar eness of personal identity will more likely be triggered by a difference in skin color from fello w test-takers and the test administrator. In addition, awareness of group identity will be ac tivated by concern with dispelling the negative stereotype of intellectual inferi ority among African-Americans. St ereotype threat activates the awareness that an individual s poor performance can be gene ralized to ones group, impairing academic performance (Cohen & Garcia, 2005). A persons awareness of others during testi ng has been used as a prime for stereotype threat in previous studies (Aronson, Lustin a, 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 supe rior or the person that descri bes the task as diagnostic of intellectual ability (S chmader & Johns, 2003; Spencer, St eele, & Quinn, 1999; Steele & Aronson, 1995). Thus, the administrator can have a significant effect on the test-taker. Consistent with Steele (1997) and Marx a nd Roman (2002) that purports providing role models can dispel stereotypes th at students may endorse, the curre nt study proposes that a test administrator that serves a pos itive role model can prevent or reduce stereotype threat among African-Americans. Group identity can allow the test-taker to rela te to the administrator as a person of the same stigmatized group that is successful academically, contradicting the stereotype. In other words, the role can elic it a sense of confidence a nd/or self-efficacy that optimizes the students performance. In addition, the ethnicity of the test administrator can influence a student of co lors level of evaluation apprehension. Walters, Shepperd, and Brown (2003) found that African-American

PAGE 31

31 undergraduates performed worse than European-Ame rican undergraduates on a standardized test when the administrator was European-Ame rican, but equal to European-American undergraduates when the administrator was Afri can-American. They proposed that students of color underperformed when th e 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 administrato r endorses stereotypes of group intellectual inferiority, evoking evaluation apprehension, distr action, and subsequent poor performance. In contrast, students of color may believe that th e African-American admini strator 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 propos ed that an African-American test administrator can also reduce stereotype threat in African-A merican 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 endor se stereotypes of group intellectual inferiority, believes that the test takers are academically comp etent, and will evaluate each test-taker on their individual performance. In addition, positive ro le 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 challe nges the plausibility of the negative stereotype and promotes that stereotypes are generalizations and do not have to apply to every member of

PAGE 32

32 the stigmatized group. Moreover, positive role m odels of the same ethnicity and/or gender further represent an attainable goal which mo tivates students to work harder and achieve academically (Lockwood & Kunda, 2002; Zirkel, 2002). The current study attempted to clarify which featur es of the test administrator are needed to reduce stereotype threat. Walters et al. (2003) found that a test ad ministrator of similar ethnicity, perceived through similar appearance, can reduce stereotype threat. Conversely, Marx and Roman (2002) purported that students must percei ve 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 colle ctive identity between the test administrator and participants. Consequently, the question of whether the other f eatures of a role model that are outlined in Marx and Roman (2002) are needed for a test administrator to se rve 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 te st administrator. Further investigation of the role of test administrator in the reduction of stereotype threat using a sample of African-American undergraduat es from a predominantly European-American (be consistent) university would be beneficial. A majority of studies that have demonstrated negative effects of stereotype threat among students of co lor have been conducted at predominantly European-American universiti es (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 whic h students of color consistently struggle academically and socially (Allen, 1992; Flower s, 2002). Demonstrating the effects of test administrator ethnicity on student performance can have great implications in the development of

PAGE 33

33 interventions and support programs for students of co lor, as well as the recruitment and retention of faculty of color, at predominan tly 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 in tellectual abilities can produce a threat that upsets the performance of women in mathematic s and students of color in academic domains (Steele, 1997). The specific research question for this study was as follows: Is similar domain identification and attainab le goal status required for an AfricanAmerican test administrator to serve as a ro le model to prevent stereotype threat among African-American students? Steeles wise schooling theory (1997) proposes that various principles of multicultural education can prevent or reduce stereot ype threat, including providing role models to students of stigmatized groups. Role models can reduce stereo type threat by dispellin g 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 me diate the effects of stereotype threat among women in the domain of mathematics (Marx & Roman, 2002). The effectiveness of a role model relies heavil y upon a students ability to relate and make social comparisons with the role model. A st udents social identity is complex, layered with ethnicity, gender, socioeconomic status, and va rious other individual an d 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 per ceive similarity in physical app earance, 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 thre at in African-American

PAGE 34

34 students. Mere similarity in physical appear ance allowed the students to identify the test administrator as a fellow member of their st igmatized group, positively impacting their academic performance under stereotype thr eat 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 prev ent stereotype threat among African-American students. Specific hypotheses for this study are as follows: A 3-way interaction between ethnicity of the ad ministrator, ethnicity of student, and role model status of administrator is expected. Specifically, ethnic similarity will prevent stereotype threat. The performance of African-American student s, the stigmatized group members, will be significantly affected by the ethnicity and role model status of the te st 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 (Mar x & Roman, 2002; Walters, Shepperd, & Brown, 2003). These hypotheses were evaluated by manipulating stereotype threat, de scribing the test as diagnostic of intellectual ability, among a sample of African-American and European-American undergraduates using African-Ame rican or European-American te st administrators and role models. Participants completed a brief test questionnaire regardi ng their test experience. Based on previous laboratory research (Schmader & Johns, 2003; Walters, Shepperd, & Brown, 2003), the questionnaire included items that assessed wh ether participants perception of the test were consistent with participants of previous laborator y investigations of stereo type threat. In other words, there were questionnaire items that assess ed participants percepti on of the test as an evaluation of group differences. In addition, an it em that required participants to estimate the administrators expectation and in terpretation of his or her performance was also included to

PAGE 35

35 evaluate the participants awareness of the test administrator. Furt hermore, an item that required participants to rate their perception of the test administrator as a role model was also included.

PAGE 36

36 CHAPTER 2 METHODS Participants and Design All participants were college students enroll ed in a large state uni versity in the south eastern United States. Participants include d 58 African-American and 54 European-American Educational Psychology students or scholarship students who partic ipated for course credit or credit towards a volunteer requirement to maintain their scholarships. Additional demographic information regarding participants is presented in Table 2-1. Participan ts were recruited for participation if they reported in a recruitmen t survey that they were African-American or European-American and had a neutral or positive at titude toward verbally-oriented classes. The recruitment survey requested demographic informa tion: race/ethnicity, age, major, SAT score, and verbal domain identificati on (Appendix A). Participants had to identify themselves in the category of European-American and African-Ame rican, activating participants awareness of group identity. The classificati on of Multiracial was provided in the recruitment survey. However, only students that iden tified themselves as solely African-American or EuropeanAmerican were eligible to participate in the st udy due to previous rese arch 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 particip ants were asked to respond to the following statement (Steele, 1995): Please i ndicate your typical at titude toward verba lly-oriented classes (literature, language, writing). Responses were answered on a Likert scale ranging from 1 (strongly dislike) to 5 (strongly like). Sim ilar to previous resear ch (Marx & Roman, 2002; Walters et al., 2003), only those who responded with a 3 or greater were eligible to participate to avoid students that dis like verbally-oriented tasks and are un likely to give maximum effort.

PAGE 37

37 Participants were also asked to provide the num ber of academic scholarships they have received to acquire anecdotal information regarding thei r academic domain identification. Participants were randomly assigned to one of four conditions in a 2 (test administra tor ethnicity: AfricanAmerican vs. European-American) 2 (test admini strator status: role model vs. non role model) 2 (student ethnicity: African -American vs. European-American ) between-subjec ts factorial design. Procedure To evoke stereotype threat va rious actions were taken. Ster eotype threat was initially activated during the recruitment phase in which st udents were asked to indi cate their ethnicity on the recruitment survey, signifyi ng that ethnicity would be a va riable in the study (Appendix A). During the experimental sessions, th e consent form served as a tri gger of stereotype threat and activated group identity, indicati ng that participants were sel ected for the study because they identified themselves as African-American or European-American (Appendix B). Stereotype threat was also activated thr ough 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 th reat. The combination of African-American and European-American pa rticipants in each session activated group identity and the subsequent social compar isons and awareness of stereotype threat. An African-American or European-American test administrator conducted each experimental session with a group of four to se ven 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 tes ting room. At the beginning of the session, the administrator officially greeted pa rticipants and described the study as an investigation of verbal reasoning abilities (Appe ndix C). To promote maximum effort in performance participants were

PAGE 38

38 told that they would receive feedback on their ve rbal strengths and weakne sses at the end of the session (Steele & Aronson, 1995). Students that are concerned about their academic performance are more likely to want to know th eir 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 di splay competence, the test administrator presented the test as a difficult exam she created and that she would be pr ovide 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 no t actually receive feedb ack on their performance. Upon completing the test, participants completed a brief questionnaire regarding their test experience (Appendix F). To test their perception of ster eotype threat, participants rate d 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 [Afr ican-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 admini strator across role model and non-role model conditions, participants rated how much they ag ree 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 pu rpose of the study and told that they would not receiv e feedback on their performance.

PAGE 39

39 Table 2-1. Demographics of sample. Total Sample African-American European-American N N N Number of Participants 112 58 54 Gender Female 90 49 41 Male 22 9 13 College Level Junior/Senior 71 35 36 Freshman/Sophomore 41 23 18 Scholarship Status Non-Scholarship 79 33 46 Scholarship Recipient 33 25 8

PAGE 40

40 CHAPTER 3 RESULTS Omnibus Analysis of Test Performance Test performance was calculated as a percen tage 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 performi ng a 2 (student ethnicity : African-American or European-American) 2 (test administrator ethnicity: African-American or European-American) 2 (test administrator status: role model or non-role model) ANCOVA on test performance, with self-reported verbal SA T score serving as the covari ate. The ANCOVA results are presented in Table 3-1. This analysis revealed a significan t effect for the covariate, F (1, 93) = 25.63, p < .001, partial = .22; no main effect for student ethnicity, F (1, 93) = 1.76, p = .19, partial = .02; no main effect for test administrator ethnicity, F (1, 93) = .52, p = .47, partial = .01; and no main effect for test administrator status, F (1, 93) = .01, p = .91, partial = .00. In addition, no interactions were found, F s < 1.16, p s > .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 administra tor across role model and non-role model c onditions, participants completed a brief questionnaire regarding thei r test experience. The questionnaire was comprised of two questions rega rding their perception of stereo type threat and one question assessing perception of the administrator. Part icipants responded usi ng a Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree) Each questionnaire item was analyzed using a 2 (student ethnicity: African-Am erican or European-American) 2 (test administrator ethnicity:

PAGE 41

41 African-American or European-American) 2 (test administrator status: role model or non-role model) ANOVA. 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 whol e, 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 thei r respective ethnic group based on their test performance. No main effect for test administrator ethnicity, F (1, 104) = .08, p = .78, partial = .00, and no main effect for test administrator status, F (1, 104) = .64, p = .43, partial = .01, were found. In addition, no interactions were found, F s < 1.84, p s > .18. Analysis of the extent to which participants were concerned that the test administrator would perceive their respective et hnic group as having less intellectua l 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 pe rceive 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 = .00, and no main effect for test administrator status, F (1, 104) = .53, p = .47, partial = .01, were found. In addi tion, no interactions were found, F s < 2.18, p s > .14. Test Administrator Stat us 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,

PAGE 42

42 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 effectivel y operationalized and presented in the role model condition. In a ddition, 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 administra tor. No main effect for test administrator ethnicity was found, F (1, 104) = .00, p = .99, partial = .00. In addition, no interactions were found, F s < 1.98, p s > .16. Table 3-1. Analysis of covarian ce for test performance Source Df F p 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 TE 1 .08 .00 .78 SE TS 1 .01 .00 .91 TE TS 1 1.16 .01 .29 SE TE 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 administrator administrator administrator administrator 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 in parentheses represent standard deviations.

PAGE 43

43 CHAPTER 4 DISCUSSION The purpose of this study was to explore factors that can prevent stereotype threat, a phenomenon in which negative stereotypes about in tellectual abilities can produce a threat that upsets the performance of women in mathematic s 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 requi red for an African-American test administrator to serve as a role model to prevent stereot ype threat among African-American students. The current study extends previous research regard ing stereotype threat by attempting to clarify which features of the test administrator are n eeded to reduce stereotype threat, just similar ethnicity (Walters et al., 2003) or similar ethnicity, domain identific ation, and attainable status of the test administrator conj ointly (Marx & Roman, 2002). African-American participants were expected to perform to the level of EuropeanAmerican participants when the test admini strator was African-American, demonstrating the inhibition of stereotype threat. Additionally, African-American pa rticipants were expected to perform better than European-American participants when the test administrator was an AfricanAmerican role model. However, no evidence of stereotype threat or it s 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 perf ormance for African-American participants. These findings were inconsistent with exp ected results that were based on previous stereotype threat research in laboratory settings (Marx & Roman, 2002; Walters et al, 2003). Conversely, these findings were cons istent with previous stereotype threat research conducted in actual test settings that found no ev idence of stereotype threat (Stricker, 19 98; Stricker & Ward, 1998; Walters, Lee, & Trapani, 2004). While th ere was evidence that African-Americans were

PAGE 44

44 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 di sruption in their performance. Although African-Americans were more likely than European-Americans to perceive the role model as know ledgeable and credible, their awareness of the role model did not lead to signifi cantly better perfor mance. Although nonsignificant differences in test performance were not unexpected, these findings are remarkable because they demonstrate that some African-American undergraduates do not experience impaired test performance even wh en they experience evaluation apprehension. While impairment of test performance is an inte gral component of stereotype threat, the current findings suggest that some African-American students experience the fear of confirming negative stereotypes, but not to th e extent that it impairs test perf ormance. Thus, future studies should continue to address under what circum stances groups who often report experiencing stereotype threat are resilient to its effects during testing. Further investigation of test environment and interpersonal factors (i.e. prio r educational experiences self-efficacy, etc.) as forms of stereotype threat prevention is warranted. It can be speculated that there was no eviden ce of stereotype threat in the current study because the environment was not like previous labor atory 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 wa s a less academic and intimidating environment than a classroom or computer lab where prev ious laboratory research was conducted (Marx & Roman, 2002; Walters et al., 2003) possibly reducing the susceptib ility to stereotype threat. There are additional explanations for the nonsignificant findings. Similar to the previous study in an authentic testing situ ation by Stricker (1998) that did not find evid ence of stereotype

PAGE 45

45 threat, participants were require d to provide their ethnicity upon completing the exam instead of prior to taking the exam. Although students were required to identify th eir 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 perceive d ethnicity to be their most sa lient 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 th e fear of evaluation and susceptibility of stereotype threat. In addition, the exam was described as an examination of personal factors involved in solving problems th at require verbal reasoning abili ties, which may have shifted participants awareness away from their group iden tity to their personal identity during the exam. Interestingly enough, African-American participan ts were more concerned than EuropeanAmerican participants about th eir performance being perceived as representative of their respective group. However, participants prov ided those responses on the questionnaire following the exam in which they were required to provide their ethni city, which may have shifted participants awareness to their group iden tity. Requiring participants to provide their ethnicity explicitly activates group identity and stereotype awarene ss while the description of the exam is more subtle. The difference in the inte nsity of the environmenta l 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 evalua tive of intellectual abilities or mathematical ability, explicitly evoking awareness of nega tive stereotypes regard ing group intellectual inferiority.

PAGE 46

46 In contrast to previous labor atory studies that involved freshmen and sophomores from introductory psychology courses (Steele, 1995; Steele & Arons on, 1995; Walters et al., 2003; Marx & Roman, 2002), a majority of the sample was comprised of upper-lever students across various college majors. Thus, the sample was sim ilar 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 sa mple was comprised of African-American and European-American students with academic sc holarships. Scholarship students may be less susceptible to stereotype threat because they ha ve earned scholarships for their superior test performance, affirming confidence in their test-t aking abilities. Students with positive selfconcepts regarding test performance may be le ss likely to fear perfor ming poorly and confirming negative stereotypes regardi ng their respective groups. Mo reover, the African-American participants were undergraduate s of a predominantly White university who have experienced success in academic settings in which there we re few African-Americans. Thus, the AfricanAmerican participants may be desensitized to a cademic experiences in which they are in the minority, subsequently reducing suscep tibility 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 ad ministrator ethnicity exclusivel y prevents stereotype threat

PAGE 47

47 (Walters et al., 2003) or whether attainable goal status and similar doma in identification also needed to prevent stereotype threat (Marx & Roman, 2002) Prior to discussing implications of this study, it is important to men tion its limitations. The most significant limitation was the diversity a nd size of the sample. Although the sample was comprised of a balanced number of African-Ame rican and European-American participants, the majority of the sample was upper-level studen ts 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-le vel 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-o riented courses. Requiring such preference possibly reduced the numbe r of male participants that are more likely to prefer mathoriented courses. Some stereotype threat st udies 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 bala nce of gender across part icipants 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 in clude the systematic i nvestigation of test administrator and student gender an d ethnicity as factors in the pr evention of stereotype threat among African-American students. Specifically, great er 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, incr easing the gender college level diversity of participants would make the sample more repres entative of the university population.

PAGE 48

48 Furthermore, data analysis of test performan ce involved self-reported verbal SAT scores as a covariate. Self-reported test scor es 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 thei r actual score (Kuncel, Crede, & T homas, 2005). It is important to note that there were no significant ethnic group di fferences in the self -reported verbal SAT scores. This study addresses various issues that are re levant 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 demons trated in previous research as a possible explanation for such disparity. These findings may be a function of the limitations in this study or an indicator that test admi nistrator ethnicity or role mode l status do not induce or reduce stereotype threat. These findings align with emerging research th at has demonstrated a lack of evidence for stereotype threat (S tricker, 1998; Stricker & Ward, 1998; Walters et al., 2004). It is possible that stereotype threat on ly occurs in laboratory settings and testing environments that evoke awareness of ethnic differences among exam inees. Future research should investigate possible explanations for the differe nces in laboratory and field study results. Further research is needed to determine which environmental factor s evoke and inhibit ster eotype threat. The lack of evidence of stereo type threat in this study demo nstrates that some AfricanAmerican students are not susceptible to stereo type threat. While pr evious 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

PAGE 49

49 needed to identify what interpersonal factors, besides ethnicity or gende r, make students less susceptible to stereotype threat. Identifying personal characteristics that decrease susceptibility to stereotype threat can enhan ce 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 performan ce 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 th at might ameliorate any adverse affects that stereotype threat may ha ve on test performance.

PAGE 50

50 APPENDIX A PARTICIPANT ELIGIBILITY SURVEY Please circle the race/ethnicity that you most identify with. White/European-American Black (non-Hispanic)/African-American Latino Asian Multiracial If you took the SAT, please circle your approximate verbal score. Did not take SAT 400 451 501 551 601 651 701 751 Please indicate the number of academic sc holarships you have ever received. _____ Please indicate your major._________________ Please indicate your typical atti tude toward verbally-oriented classes (literat ure, language, writing). Strongly Dislike Neutral Strongly Like 1 2 3 4 5

PAGE 51

51 APPENDIX B CONSENT FORM You are invited to participate in a research study that investig ates personal factors involved in performance on problems requiring reading an d 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 agre ed to participate on a voluntary basis and have expressed a neutral or positive at titude 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 Ps ychology department, is conducting the study. Dr. Tracy Linderholm, Assistant Professor of Educati onal Psychology, University of Florida, and Dr. Nancy Waldron, Associate Professor of Sc hool Psychology, are supervising this study. Background Information: The purpose of this study is to understand pers onal factors that im pact performance on problems that require reading and verbal reasoni ng abilities. The results of this study will have educational implications in te rms of providing guidance for sta ndardized test performance. Procedures: You will be asked to complete a 17-item test of verbal problems. Following the test you will be asked to complete a short su rvey 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 participa tion in this study. The only possible risk of participating in this study is mental fatigue. The probability of this risks occurrence is extremely low. Compensation: Scholars: Participation in research earns course or scholarship credit. Participants can earn volunteer credits toward maintaining their scholar ships as dictated by th e Office of Admissions. or Education course students: Participants can ea rn extra credits determined by the individual course instructor. Instru ctors will be notified that you have participated. Voluntary Nature of the Study: Your decision whether or not to participate wi ll not affect your curren t or future relations with the University of Florida. If you decide to participate, yo u are free to withdraw at any time without affecting those relati onships and without forfeiting the compensation (course or scholarship credit) that was agreed upon. Confidentiality: 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 ca n be accessed only by the

PAGE 52

52 researcher. When the study is completed and the data have been analyzed, the list will be destroyed. In addition, the experi mental results will be analyzed in terms of averages across participants rather than in terms of individua l performances. In any form of report we might publish, we will not include any information that will make it possible to identify a research participant. Contacts and Questions: The researcher conducting this study is Ra shida Brown, a doctoral student. Dr. Tracy Linderholm, Assistant Professor of Educationa l 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 ques tions later, please contact the experimenter, Rashida Brown, Educational Psychology, (352) 246-5100 or Dr. Tracy Linderholm, Educational Psychology, (352) 392-0723 ext. 241 or Dr. Na ncy 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 ri ghts as a research participant, please contact UFIRB office, P.O. Box 112250, University of Florida, Ga inesville, 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 answ ers. I consent to pa rticipate in the study. Signature of Participant __________________________ Date _______________ Signature of Investigator _________________________ Date _______________

PAGE 53

53 APPENDIX C TEST ADMINISTRATOR SCRIPT FO R NON-ROLE MODEL CONDITION Hello, My name is ________________. Thank you for your particip ation. The purpose of this study is to investigate various personal factors involve d in performance on probl ems requiring reading and verbal reasoning abilities. Y ou are asked to complete a 17-item test of verbal problems in a format identical to the SAT exam and a short surv ey 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 i nvolved 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 an alysis of your verbal ability. Please complete each question. Since this is an analysis of your ve rbal 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 fo r your participation. Are there any questions?

PAGE 54

54 APPENDIX D TEST ADMINISTRATOR SCRIPT FOR ROLE MODEL CONDITION Hello, 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 disse rtation 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 co mplete 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 genui ne test of your verbal abilities and limitations so that I might better understand the factors i nvolved 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 strength s and weaknesses in verbal problem solving. Although you can expect not to get many of the questions correct, pleas e 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 th ere any questions?

PAGE 55

55 APPENDIX E STUDY TEST QUESTIONS Directions : Each sentence below has one or two bla nks, each blank indica ting that something has been omitted. Beneath the sentence are five 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 psychiat ry does not deny the contributi ng roles of psychological factors in mental illness, but that these factors may act as a catalyst on existi ng physiological conditions and _____ such illnesses. A. disguise B. impede C. constrain D. precipitate E. consummate 2. During periods of social and cultural st ability, many art academies are so firmly controlled by _____ that all real creative wo rk 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 literatu re, skillfully used, could make it. A. circumlocution . literal B. clich . 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 da ta that were inconsistent with ot her findings, in this particular experiment she let a number of _____ results slip by.

PAGE 56

56 A. inaccurate B. verifiable C. redundant D. salient E. anomalous 6. Czannes delicate watercolor sketches often served as _____ of a subject, a way of gathering fuller knowledge before the artists fina l engagement of the subject 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 fa r 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 standard s of decent conduct mandated by company policy if _____ compliance with instructio ns 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

PAGE 57

57 Directions : In each of the following que stions, a related pair of wo rds or phrases is followed by five lettered pairs of words or phr ases. 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

PAGE 58

58 Directions : The passage below is followed by questions based on its content. After reading the passage, choose the best answer to each questi on. Answer all questions following the passage on the basis of what is stated or implied in that passage. The historian Frederick J. Turn er wrote in the 1890s that th e agrarian discontent that had been developing steadily in the United States since about 1870 had been precipitated by the closing of the internal frontierthat is, the depl etion of available new land needed for further expansion of the American farming system. Not onl y was Turners thesis influential at the time, it was later adapted and elaborated by ot her 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 1890s, when agrarian discontent had become most acute, 1,100,000 new farms were se ttled, 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 farmi ng 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 pr actices 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 states men to the presumed disappearance of the American frontier helped to obscure the grea t importance of changes in the conditions and consequences of international trade that occu rred during the second half of the nineteenth century. In 1869 the Suez Canal was opened and the first transcontinental ra ilroad in the United Sates was completed. An extensive network of telegraph and tele phone 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 ag ricultural 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 we re joined with one another and with the countries of Europe into an interdependent market system. As a consequence, agrarian de pressions no longer were local or national in scope, and they struck se veral nations whose inte rnal 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 unint errupted decline in the prices of American agricultural products on foreign ma rkets. Those staple-growing farmers in the United States who exhibited the greatest discontent were those w ho had become most depe ndent on foreign markets for sale of their products. In sofar 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 so. 14. According to the author, changes in the cond itions 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 ne twork for shipping agricultural products D. extension of agrarian depr essions beyond national boundaries E. emphasis on the importance of market forces in determining the process of agricultural products

PAGE 59

59 15. The author implies that the change in that state of the American farmers morale during the latter part of the nineteenth century was traceab le to the American farmers increasing perception that the: A. costs of cultivating the land were prohibitive within the United States B. development of the first transcontinental rail road 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 its expansion D. prices of American agricultural products were deteriorating es pecially rapidly on domestic markets E. proceeds from the sales of American agri cultural products on foreign markets were unsatisfactory 16. According to the passage, which of the following occurred prior to 1890? A. Frederick J. Turners 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 States C. The manufacturers of technologically advanced agricultural machinery rapidly increased their marketing efforts D. Direct lines of communicati on were constructed between the United States and South America E. Technological advances made it fruitful to farm extensively on a mechanized basis 17. The authors argument implies that, compared to the yearly price changes that actually occurred on foreign agricultura l 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

PAGE 60

60 APPENDIX F 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-American s/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

PAGE 61

61 LIST OF REFERENCES Allen, W. R. (1992). The color of success: African American college student outcomes at predominantly white and historically bl ack public college a nd universities. Harvard Educational Review, 64 (1) 26. Ambady, N., Paik, S. K., Steele, J., Owen-Smith, A., Mitchell, J. P. (2004). Deflecting negative self-relevant stereotype activatio n: The effects of individuation. Journal of Experimental Psychology, 40, 401. Ambady, N., Shih, M., Kim, A., and Pittinsky, T. L. (2001). Stereotype susceptibility in children: Effects of identity activ ation on quantitative performance. Psychological Science, 12, 385. Aronson, J. (2002). Stereotype threat: Conten ding and coping with unnerving expectations In J. Aronson (Ed.), Improving academic achievement: Impac t of psychological factors on education (pp. 279) New York, NY: Academic Press. Aronson, J., Fried, C. B., and Good, C. (2002). Re ducing the effects of st ereotype threat on African American college students by shaping theories of intelligence. Journal of Experimental Social Psychology, 38 : 113. Aronson, J., Lustina, M. J., Good, C., Keough, K. Steele, C. M., & Brown, J. (1999). When white men cant do math: Necessary and sufficient factors in stereotype threat. Journal of Experiment al Social Psychology, 35, 29. 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 targets perspective (83-103).. New York, NY: Academic Press. Baddeley, A. (1996). Explori ng the central executive. The Quarterly Journal of Experimental Psychology, 49A (1), 5. Ben-Zeev, T., Fein, S., & Inzlicht, M. (2005). Arousal and stereotype threat. Journal of Experimental Social Psychology, 41, 174. Blascovich, J., Spencer, S. J., Quinn, D., & Steel e, C. M. (2001). African Americans and high blood pressure: The role of stereotype threat. Psychological Science, 12 225. Brown, L. M., & Dobbins, H. (2004). Students of color and European American students stigma-relevant perceptions of university instructors. Journal of Social Issues, 60( 1), 157. Cohen, G. L., & Garcia, J. (2005) I am us: Negative stereo types as collective threats. Journal of Personality and Social Psychology, 89(4), 566. Croizet, J., Desprs, G., Gauzins, M., Huguet, P., Leyens, J., & Mot, A. (2004). Stereotype

PAGE 62

62 threat undermines intellectual performan ce by triggering a disruptive mental load. Personality and Social Psychology Bulletin, 30(6), 721. Croizet, J., Dsert, M., Dutrv is, 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 of Education, 4 295. Demo, D. H., & Parker, K. D. (1987). A cademic achievement and self-esteem among African-American and White college students. Journal of Social Psychology 127 345 355. Educational Testing Service. (2003). Sex, race, ethnicity, and performance on the GRE General Test 2003-2004. Retrieved March 1, 2005, from http://ftp.ets.org/pub / gre/994950.pdf 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. Good, C., Aronson, J., & Inzlicht, M. (2003). Im proving adolescents standardized test performance: An intervention to redu ce the effects of stereotype threat. Applied Developmental Psychology, 24 645. 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. Herrnstein, R. J., & Murray, C. (1994). The bell curve. Intelligence and class structure in American life. New York, NY: Free Press. Hess, T. M., Auman, C., Colcombe, S. J., & Rahha l, T. A. (2003). The impact of stereotype threat on age differences in memory performance. Journal of Gerontology, 58 3. Inzlicht, M., & Ben-Zeev, T. (2000). A threaten ing intellectual environment: Why women are susceptible to experience problem-sol ving deficits in the presence of men. Psychological Science, 11 365. Jorna, P. (1992). Spectral analysis of heart rate and psychological state: A review of its validity as workload index. Biological Psychology, 34, 237. Keller, J., & Dauenheimer, D. (2003 ). Stereotype threat in the cl assroom: Dejection mediates the disrupting threat effecton wo men's math performance. Personality and Social Psychology Bulletin, 29 371. Kozol, J. (1991). Savage inequalities: Childre n in Americas schools. New York, NY: Harper Perennial.

PAGE 63

63 Kuncel, N. R., Crede, M. & Thomas, L. L. ( 2005). The validity of self-reported grade point averages, class ranks, and test scores: A me ta-analysis and review of the literature. Review of Educational Research, 75, 63. Lockwood, P., & Kunda, Z. (2002). Motivation by pos itive or negative role models: Regulatory focus determines who will best inspire us. Journal of Personality and Social Psychology, 83(4), 854. Markus, H. R., & Kitayama, S. (1994). The cult ural construction of self and emotion: implications for social behavior. In S. Kitayama & H. R. Markus (Eds.). Emotion and culture: Empirical studie s of mutual influence (pp. 89). Washington, DC: American Psychological Association. Marx, D. M., & Roman, J. S. (2002). Female role models: Protecting womens math test performance. Personality and Social Psychology Bulletin, 28(9) 1183. McKay, P. F., Doverspike, D., Bowen-Hilton, D. McKay, Q. D. (2003). The effects of demographic variables and ster eotype threat on black/white differences in cognitive ability test performance. Journal of Business and Psychology, 18 1. Nieto, S. (2004). Affirming diversity: The so ciopolitical context of mu lticultural education (4th edition). New York, NY: Longma n Publishing Group. O'Brien, L. T., & Crandall, C. S. (2003). Stereot ype threat and arousal: E ffects on women's math performance. Personality and Social Ps ychology Bulletin, 29 782. Ogbu, J. U. (1978). Minority education and caste: The Am erican 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. Osborne, J. W. (2001). Testing stereotype th reat: Does anxiety explain race and sex differences in achievement? Contemporary Educational Psychology, 26 291. Quinn, D. M., & Spencer, S. J. (2001). The interference of stereo type threat with womens generation of mathemati cal problem-solving strategies. Journal of Social Issues, 57 55. Sathy, V., Barbuti, S., Mattern, K. (2006). The new SAT and trends in test performance. Retrieved October 14, 2006, from College Board Web site: http://www.collegeboard com/research/abstract/150063.html Schmader, T. (2002). Gender identification moderate s stereotype threat ef fects on womens math performance. Journal of Experimental Social Psychology, 38, 194.

PAGE 64

64 Schmader, T., & Johns, M. (2003). Converging evidence that stereotype threat reduces working memory capacity. Journal of Personality and Social Psychology, 85 440. Sekaquaptewa, D., & Tompson, M. (2003). Solo st atus, stereotype thr eat and performance expectancies: Their effects on women's performance. Journal of Experimental Social Psychology, 39 68. Shih, M., Pittinsky, T. L., & Ambady, N. (1999). Stereo type susceptibility: Shifts in quantitative performance from sociocultural identification. Psychological Science, 10 81 Smith, J. L. (2004). Understanding the proc ess of stereotype threat: A review of mediational variables and new performance goal directions. Educational Psychology Review, 16 (3), 177). Smith, C. E., & Hopkins, R. (2004). Mitigating the impact of stereotypes on academic performance: The effects of cultural identi ty and attributions for success among African American college students. The Western Journal of Black Studies, (28)1, 312. Smith, J. L. & White, P. H. (2001). Development of the domain identifica tion measure: a tool for investigating stereoty pe threat effects. Educational and Psychological Measurement, 61 1040. Spencer, S. J., Steele, C. M., & Quinn, D. M. (1999). Stereotype threat and womens math performance. Journal of Experimental Social Psychology, 35 4. 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 stereotype s shape intellect ual identity and performance. In J. Banks & C. Banks (Eds.), Handbook of research on multicultural education (2nd edition) (pp. 682) San Francisco, CA: Jossey-Bass. Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectua l identity and performance. American Psychologist, 52 613. Steele, C. M., & Aronson, J. (1995) Stereotype threat and the in tellectual test pe rformance of African Americans. Journal of Personality and Social Psychology, 69 797. Steele, C. M., & Aronson, J. (1998). Stereot ype threat and the test performance of academically successful African Americans In C. Jencks & M. Phillips (Eds.), The black-white test score gap (pp. 401). Washington, D.C. : Brookings Institution Press. Stone, J., Lynch, C. I., Sjomeling, M., & Darley, J. M. (1999). Stereotype threat effects on black and white athletic performance. Journal of Personality and Social Psychology, 77 1213 1227.

PAGE 65

65 Stricker, L. J. (1998). Inquiring about examinees ethnicity and sex: Effects on AP Calculus AB Examination performance (College Board Rep. No. 98; ETS RR 98). New York, NY: College Board. Stricker, L. J., & Ward, W. C. (1998). Inquiring about examinees ethni city and sex: Effects on computerized placement tests performance (College Board Rep. No 98-02; ETS RR 9809). Princeton, NJ: Educatio nal Testing Service. Tajfel, H., & Turner, J. C. (1986). The social id entity theory of intergroup behavior. In S. Worchel & W. G. Austin (Eds.). The psychology of in tergroup relations (pp. 7). Chicago: Nelson Hall. Taylor, E., & Antony, J. S. (2000) Stereotype threat reduction a nd wise schooling: Towards the successful socialization of African American doctora l students in education. The Journal of Negro Education, 69 (3), 184. Turner, M. L., & Engle, R. W. (1989) Is working memory task dependent? Journal of Memory and Language, 28 127. Urdan, T. C., & Maehr, M. L. (1995). Beyond a tw o-goal theory of motiva tion and achievement: A case for social goals. Review of Educational Research 65, 213. Walters, A. M., Lee, S., & Trapani, C. (2004). Stereotype Threat, the Te st-Center Environment, and Performance on the GRE General Test (College Board Rep. No 01-03R; ETS RR04-37). 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 w ith questionnaire measures. Journal of Personality and Social Psychology, 72 262. Worchel, S., & Coutant D. (2001). It takes two to tango: Relating group identity to individual identity within the framework of group deve lopment. In M. A. Hogg & R. S. Tindale (Eds.), Blackwell handbook of social psychology: Group processes (pp. 461). Malden, MA: Oxford Blackwell Publishing. Zirkel, S. (2002.) Is there a place for me? Ro le models and academic identity among white students and students of color. Teachers College Record, 104(2), 357.

PAGE 66

66 BIOGRAPHICAL SKETCH Rashida Williams Brown was born on December 19, 1979 in Miami, Florida. The youngest of three children, she gr ew 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 Flor ida in 2001 and 2005, respec tively. She earned a Ph.D. in School Psychology in 2007.


Permanent Link: http://ufdc.ufl.edu/UFE0019520/00001

Material Information

Title: Exception to the Rule: Can Test Administrators Prevent Stereotype Threat?
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0019520:00001

Permanent Link: http://ufdc.ufl.edu/UFE0019520/00001

Material Information

Title: Exception to the Rule: Can Test Administrators Prevent Stereotype Threat?
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0019520:00001


This item has the following downloads:


Full Text





EXCEPTION TO THE RULE:
CAN TEST ADMINISTRATORS PREVENT STEREOTYPE THREAT?





















By

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


2007





































O 2007 Rashida Williams Brown


































To all who nurtured my faith, dreams, and scholarship









ACKNOWLEDGE1VENTS

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

much.












TABLE OF CONTENTS


page

ACKNOWLEDGEMENT S ............ ..... ._ ...............4....


LIST OF TABLES ............ ..... .__ ...............7...


AB S TRAC T ......_ ................. ............_........8


CHAPTER


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................


APPENDIX


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


Table page

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?

By

Rashida Williams Brown

May 2007

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

stereotype threat.









CHAPTER 1
INTTRODUCTION

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

academic performance.

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

threat.

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

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).

Working Memory

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

generally supportive.

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

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.

Effort

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

situation.

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

threat.

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.









Proposed Model

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

dimensions.

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
African-American students?

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
stereotype threat.

* 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,
2003).

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.









CHAPTER 2
METHOD S

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

design.

Procedure

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
Gender
Female
Male
College Level
Junior/Senior
Freshman/Sophomore
Scholarship Status
Non-Scholarship
Scholarship Recipient


Total Sample African-American
N N
112 58


European-American
N
54

41
13









CHAPTER 3
RESULTS

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

model) ANOVA.

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 >

.16.

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









CHAPTER 4
DISCUSSION

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

inferiority .









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-

oriented courses.

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

scores.

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.










APPENDIX A
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
Latino
Asian
Multiracial



If you took the SAT, please circle your approximate verbal score.

Did not take SAT
400-450
451-500
501-550
551-600
601-650
651-700
701-750
751-800


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,
writing).

Strongly Dislike Neutral Strongly Like

1 2 3 4 5









APPENDIX B
CONSENT FORM

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.

Background Information:
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.

Procedures:
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
low.

Compensation:
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.

Confidentiality:
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
participant.

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









APPENDIX C
TEST ADMINISTRATOR SCRIPT FOR NON-ROLE MODEL CONDITION

Hello,

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?









APPENDIX D
TEST ADMINISTRATOR SCRIPT FOR ROLE MODEL CONDITION

Hello,

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?










APPENDIX E
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.

A. disguise
B. impede
C. constrain
D. precipitate
E. consummate

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.











A. inaccurate
B. verifiable
C. redundant
D. salient
E. anomalous

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
so.

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
products










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
that the:

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
its expansion
D. prices of American agricultural products were deteriorating especially rapidly on
domestic markets
E. proceeds from the sales of American agricultural products on foreign markets were
unsatisfactory

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
States
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
America
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









APPENDIX F
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),
157-174.

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-
355.

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/
gre/994950.pdf

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
Perennial.










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,
83(4), 854-864.

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
Psychological Association.

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.
com/research/a~bstract/1~50063.html

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,
61, 1040-1057.

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
Press.

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-
1227.










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.









BIOGRAPHICAL SKETCH

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.