social equity theory and racial-ethnic achievement gaps€¦ · social equity theory and...

Social Equity Theory and Racial-Ethnic Achievement Gaps

Clark McKownRush University Medical Center

In the United States, racial-ethnic differences on tests of school readiness and academic achievement continue.A complete understanding of the origins of racial-ethnic achievement gaps is still lacking. This articledescribes social equity theory (SET), which proposes that racial-ethnic achievement gaps originate from twokinds of social process, direct and signal influences, that these two kinds of processes operate across develop-mental contexts, and that the kind of influence and the setting in which they are enacted change with age.Evidence supporting each of SET’s key propositions is discussed in the context of a critical review of researchon the Black–White achievement gap. Specific developmental hypotheses derived from SET are described,along with proposed standards of evidence for testing those hypotheses.

This article offers an account of the varied socialprocesses that contribute to mean differences in testscores between children from different racial-ethnicgroups, spanning preschool to high school (Jencks& Phillips, 1998; Lee, 2002; Lee & Burkam, 2002). Inthe United States, on measures of school readinessand academic achievement, Asian Americansachieve higher average scores than White students,who in turn achieve higher average scores thantheir Black and Latino peers (Jencks & Phillips,1998; Lee, 2002; Reardon & Galindo, 2009). Becauseit has received a great deal of attention, the Black–White achievement gap is used to illustrate themajor points of this article. The broad goal of thisarticle, however, is to propose a model that appliesto a variety of racial-ethnic and other achievementgaps.

Racial-ethnic achievement gaps are substantiallife-span phenomena. By all accounts, the magni-tude of the Black–White achievement gap is consid-erable, ranging from .5 to 1.0 SD, depending on thesample and the measure (Jencks & Phillips, 1998;Lee, 2002; Reardon & Galindo, 2009; Reardon &Robinson, 2007; Vanneman, Hamilton, Anderson, &Rahman, 2009). The Black–White achievement gapaffects individuals and the generation to which theybelong, beginning in early childhood and spanning

all educational levels (Farkas, 2003). In terms ofschool readiness, research has consistently demon-strated that prior to school entry, Black studentsachieve lower average scores than White students(Brooks-Gunn, Klebanov, & Duncan, 1996; Duncan,Brooks-Gunn, & Klebanov, 1994; Lee & Burkam,2002). Over the course of elementary school, theBlack–White achievement gap appears to grow(Farkas, 2003; Fryer & Levitt, 2004, 2005; Phillips,Crouse, & Ralph, 1998).

The Black–White achievement gap is a highlyconsequential social problem. School readiness andacademic achievement are associated with the kindsof jobs and wages people are able to secure. Racial-ethnic achievement gaps that begin at school entryand persist through school completion thus caninfluence racial-ethnic gaps in socioeconomic status(SES) across the life span (e.g., Levin, 2009; Reardon& Robinson, 2007). In turn, SES is robustly associ-ated with health (Adler, Boyce, Chesney, & Cohen,1994; Levin, 2009). Furthermore, the health of anydemocratic society is predicated on the ability of itspopulation to make informed choices at the ballotbox. When large segments of the population areinadequately educated, democracy’s health is atrisk.

As with any social problem, how policy makers,practitioners, and the public formulate the Black–White achievement gap’s cause will guide what isdone, and what is not done, to solve the problem(Humphreys & Rappaport, 1993; McKown, 2005;

This study was made possible by a William T. Grant Founda-tion Scholar’s Award to Clark McKown. Thanks to Laura Gum-biner, Anne Gregory, Stephen Quintana, Michael Strambler, andRhona Weinstein for their helpful comments on an earlier draftof this article.

Correspondence concerning this article should be addressed toClark McKown, Rush NeuroBehavioral Center, 4711 Golf Road,Suite 1100, Skokie, IL 60076. Electronic mail may be sent [email protected].

© 2012 The AuthorChild Development © 2012 Society for Research in Child Development, Inc.All rights reserved. 0009-3920/2013/8404-0002DOI: 10.1111/cdev.12033

Child Development, July/August 2013, Volume 84, Number 4, Pages 1120–1136

Seidman, 1983). Almost all accounts of racial-ethnicachievement gaps acknowledge the influence ofmultiple processes in multiple settings (e.g., Boykin,1986; Garcia-Coll, 1990; Weinstein, 2002; Weinstein& McKown, 1998). Nevertheless, much prior workon the Black–White achievement gap has focusedon specific processes operating in a limited range ofcontexts. For example, some have argued that racialdifferences in genetic endowment cause the gap(Jensen, 1969; Rushton & Jensen, 2005). Others havefocused on the contribution of SES and family fac-tors (Brooks-Gunn, Klebanov, Smith, Duncan, &Lee, 2003), cultural values (Ogbu, 2002; Thernstrom& Thernstrom, 2002), academic stereotypes (Steele& Aronson, 1995), and degree of match betweenhome and school environments (Brice-Heath, 1983;Tharp, 1989).

More focused accounts are sometimes criticizedbecause they do not account for the entire racial-ethnic achievement gap, even though they are ofteninterpreted as doing so. For example, in a criticalreview, Sackett, Hardison, and Cullen (2004)acknowledged that one social process—stereotypethreat—can depress Black students’ test scores.However, they strongly questioned common inter-pretations of the research that stereotype threat isthe sole cause of the Black–White test score gap,and that if it were eliminated, the gap would there-fore disappear.

That theory and research have developed in atargeted way is understandable—evaluating multi-factorial accounts of the gap’s origin is challenging.The conceptual challenge is to broaden inquirywithout mounting atheoretical “fishing expeditions”that provide few insights. The practical challenge isto muster resources to meaningfully evaluate thecombined influences of multiple social processesacross key developmental contexts. What is neededis a theory broad and flexible enough to account forvaried sources of social influence, yet specificenough to provide a parsimonious account of theachievement gap.

The purpose of this article is to describe socialequity theory (SET), a novel account of social pro-cesses that together give rise to racial-ethnicachievement gaps. SET provides a heuristic fororganizing and integrating research on racial-ethnicachievement gaps. SET is proposed as a theorybecause: (a) it makes specific commitments aboutthe social processes that are relevant to understand-ing the achievement gap, (b) it offers a causalexplanatory framework to explain the racial-ethnicachievement gap, and (c) it generates specific, falsi-fiable hypotheses.

SET Defined

SET describes social processes that contribute toracial-ethnic achievement gaps. SET explains meandifferences in achievement by members of differentracial-ethnic groups. There is, of course, consider-able within-group variability in academic readinessand achievement. Within-group and between-groupdifferences in academic readiness and achievementare distinct. SET does not address the causes ofindividual variability in achievement. Rather, itexplains factors that together create between-groupdifferences in school readiness and achievement.

In the context of SET, the term social processesrefers to transactions between individuals, includingverbal and nonverbal communication directed fromone person to another. The term social processes alsorefers to communications between individuals andsocial settings, in which an event or characteristic ofthe setting—apart from interpersonal interactions—communicates something of social consequence. Forexample, a poster depicting a civil rights event on aclassroom wall may communicate something ofsocial consequence without involving an interper-sonal communicative transaction.

SET adopts an ecological perspective (Bronfen-brenner, 1977; Kelly, 1987; Spencer, 1999; Weinstein,2002) to formulate what social processes in whatcontexts create and maintain racial-ethnic achieve-ment gaps. SET includes four propositions aboutthe origins of racial-ethnic achievement gaps:

1. Two classes of social process influence racial-ethnic achievement gaps. Direct influences aresocial processes that support achievement.Direct influences contribute to the racial-ethnicachievement gap when they are distributed dif-ferently to people from different racial-ethnicgroups. Signal influences are cues that commu-nicate negative expectations about a child’sracial-ethnic group. When children from nega-tively stereotyped groups detect such cues, thiscan erode achievement.

2. Signal influences depend on children’s abilityto detect cues signaling a stereotyped expecta-tion, and this ability increases significantly dur-ing the elementary grades.

3. Social processes affecting the achievement gapoperate across a limited range of key develop-mental settings, and the relevant settingschange lawfully with age.

4. Together, relevant direct and signal influencesacross developmental contexts account for theachievement gap.

Social Equity Theory 1121

From these propositions, it is possible to developspecific and falsifiable developmental hypothesesabout the origins of racial-ethnic achievement gaps.What follows is a review of evidence supportingthe existence of direct and signal influences on theachievement gap, their combined influence, and thecontexts in which they influence racial-ethnicachievement gaps. The review focuses on researchinto the Black–White achievement gap. A sub-sequent section articulates specific hypothesesuniquely derived from SET. The article finisheswith proposed standards of evidence to guidefuture work on SET and its component parts.

Evidence in Support of SET Propositions

Direct Influence on the Achievement Gap

Direct influences, defined. Direct influences aresocial processes that promote academic achieve-ment similarly for all children in all racial-ethnicgroups. SET proposes that direct influences contrib-ute to racial-ethnic achievement gaps when they aresystematically and unequally distributed to mem-bers of different racial-ethnic groups. Direct influ-ences on racial-ethnic achievement gaps may unfoldat home, in school, with peers, and in neighbor-hoods.

Home. In the home, particularly for young chil-dren, parent–child interactions and the relationshipswithin which they take place are important contextsfor children’s development. A substantial body ofresearch suggests that the characteristics and qual-ity of parent–child interactions influences preaca-demic and academic outcomes. For example,Baumrind (1966; see also Baumrind & Black, 1967),focusing largely on White middle-class samples,examined the relation between parenting styles anda variety of life outcomes among children. Theyexamined two dimensions of parenting style: sup-portiveness and demandingness. Supportivenessreferred to the level of expressed parental love, nur-turance, and responsiveness. Demandingnessreferred to the clarity of rules and the firmness andfairness of parental discipline practices. They foundthat when parents provided a combination of highsupport and high demand—what they called“authoritative” parenting—children had better aver-age academic, social, and emotional outcomes.

There is some evidence that what normative par-enting is varies from one racial-ethnic group to thenext. Mandara and Murray (2002) found thatBaumrind’s (1966) parenting styles typology charac-terized Black parenting, but that Black authoritative

parenting included greater demand and less com-promise than White authoritative parenting. Inaddition, the relation between parenting and aca-demic outcomes can differ for different groups. Forexample, at least two independent research groupshave found that among Black children, the higherthe overall neighborhood distress, the more strictparenting is associated with positive academic out-comes (Baldwin, Baldwin, & Cole, 1990; Gonzales,Cauce, Friedman, & Mason, 1996).

Prior research thus suggests that the same par-enting practices do not always promote the sameacademic outcomes for children from differentracial-ethnic groups, leaving open the question ofwhether parenting is a direct influence. The litera-ture cited above suggests that what is optimal par-enting for the development of academic outcomesmay be different for children from different racial-ethnic groups and in different contexts. However,the same research suggests that for children fromall racial-ethnic groups, achievement-supportingparenting—in whatever form that takes for thechild’s racial-ethnic group and context—plays a keyrole in the development of children’s preacademicand academic skills.

For optimal parenting practices to exert a directinfluence on the Black–White achievement gap, itmust be present at different average levels in Blackand White homes. In an important review article,Brooks-Gunn and Markman (2005) reported thatthere are racial-ethnic differences in parenting prac-tices. They reported that, compared to White par-ents, Black parents on average engaged lessfrequent warm and sensitive parenting and morefrequent negative and intrusive parenting. It isimportant to note that effect sizes were small, indi-cating a high degree of overlap in the distributionsof parenting behavior. In addition, they reporteddifferences in the frequency and richness of lan-guage exposure and availability of books and otherstimulating materials, with Black parents on aver-age engaging in less frequent conversation andhaving fewer books and other media available.

Parenting is thus associated with preacademicand academic skill development, and optimalparenting is, on average, more available to Whitechildren than to Black children. That does not meanthat these factors necessarily explain the Black–White achievement gap. It is possible, for example,that parenting and achievement gaps both arisefrom a third variable, such as Black–White socioeco-nomic differences. If this were true, in empiricalmodels with SES and parenting as independentvariables and academic score as the dependent vari-

1122 McKown

able, SES, but not parenting, would be associatedwith academic outcomes. Existing evidence runscontrary to this conclusion. Examining the ECLS–Kdata, when Lee and Burkam (2002) controlled forSES, the magnitude of the racial-ethnic school readi-ness gap was reduced, but not eliminated. Simi-larly, in secondary analyses of several large datasets, Magnuson and Duncan (2006) found that SESaccounts for some, but not all, of the Black–Whitetest score gap among young children. Furthermore,in middle childhood and adolescence, accountingfor SES reduces, but does not eliminate the Black–White test score gap (Phillips, Brooks-Gunn, Duncan,Klebanov, & Crane, 1998).

When SES and racial-ethnic differences in parent-ing practices are accounted for, the size of theschool readiness gap is dramatically reduced. Forexample, two studies found that among preschool-ers, maternal warmth and engagement account formuch of the Black–White gap that remained whenSES was accounted for (Brooks-Gunn et al., 1996,2003). Another study found that when controllingfor SES, all the residual early childhood Black–White reading readiness gap and much of the resid-ual math readiness gap was accounted for byfamily factors such as the frequency of reading (Lee& Burkam, 2002).

Parenting also helps explain the Black–Whiteachievement gap in older children. For example,Mandara, Varner, Greene, and Richman (2009)examined the relation between parenting practicesand the Black–White adolescent achievement gap.They examined multiple parenting practices, suchas involving adolescents in decision making, paren-tal monitoring, number of household chores, schoolorientation, and maternal warmth. They reportedthat: (a) White parents generally engage in moreachievement-promoting practices than Black par-ents, (b) this difference is largely accounted for byracial-ethnic differences in SES, and (c) accountingfor multiple parenting practices that promoteachievement explains the entire adolescent achieve-ment gap (Mandara et al., 2009).

In summary, optimal parenting practices: (a)promote preacademic and academic outcomes for allchildren; (b) are systematically and unequally distrib-uted by racial-ethnic group; (c) are not an artifact ofone critical potential confound, SES; and (d) accountfor much of the achievement gap that remains whenSES is accounted for. Taken together, these findingssuggest that parenting serves as a direct influence onthe Black–White achievement gap.

School. The quality of instruction and the qualityof student–teacher relationships may act as direct

influences on racial-ethnic achievement gaps. Eachof these school factors affects achievement similarlyfor students from all racial-ethnic groups. In addi-tion, high-quality instruction and positive student–teacher relationships are more available for Whitethan for Black students.

These direct influences can operate both betweenand within schools (Hanushek & Rivkin, 2009;Page, Murnane & Willett, 2008). In terms ofbetween-school influences, for example, comparedto White students, Black students attend schools inwhich instructional quality and teacher skill are, onaverage, lower (Clotfelter, Ladd, & Vigdor, 2004).Similarly, Lee and Burkam (2002) reported that atschool entry, children from different racial-ethnicgroups benefited from different levels of teacherexperience, with African American children beingassigned to less experienced teachers than theirWhite and Asian peers. Thus, systematic racial-ethnic differences in the availability of experiencedteachers and high-quality instruction betweenschools may exert a direct influence on the achieve-ment gap.

Within-school processes may also exert a directinfluence on the achievement gap. One importantwithin-school process that can contribute to theachievement gap is mean Black–White differencesin the quality of instruction. For example, through-out the primary and secondary grades, Black chil-dren are more often assigned to lower tracks thantheir White peers (Entwisle, Alexander, & Olson,1997; Lucas & Berends, 2002). Children in differenttracks are in turn exposed to different qualityinstruction, with students in higher tracks exposedto interesting and challenging material. Students inlower tracks, in contrast, are exposed to work thatis repetitive, uninteresting, and focused on behav-ioral control (Boykin, 1986; Oakes, 2005; Weinstein,2002). The quality of curriculum is thus a directinfluence because it is associated with academicoutcomes and it varies systematically by racial-ethnic group.

Black–White mean differences in the quality ofrelationships between teachers and students canalso contribute to the achievement gap. The level ofcloseness and caring children and adolescents expe-rience with their teachers influences academicoutcomes (Gregory & Weinstein, 2004; Hamre &Pianta, 2005; Weinstein, 2002). On average, Whitestudents enjoy closer and more caring relationshipwith their teachers than Black students. For exam-ple, in a study that included 197 preschool and kin-dergarten teachers, Saft and Pianta (2001) foundthat White students enjoy less conflict with and


dependency on their teachers than their Black peers.Similarly, in a study that included 607 academicallyat-risk first graders, teachers reported higher qualityrelationships with White compared to Black stu-dents (Hughes, Gleason, & Zhang, 2005). Othershave found that teachers on average expect more ofWhite students than Black students with similarrecords of achievement (Baron, Tom, & Cooper,1985; Ferguson, 2008; McKown & Weinstein, 2008).Thus, there is evidence that the quality of teacher–student relationships may be a direct influence onthe achievement gap: Those relationships are associ-ated with academic outcomes and are, on average,higher quality for White students than their Blackpeers.

Educational field trials have provided evidencethat equalizing instructional quality can reduce thegap. Success for All, a literacy curriculum that com-bines intensive focus on phonics and adherence toparticular instructional practices, has been shown toreduce the achievement gap. Furthermore, thelonger students are exposed to the curriculum, thegreater the reduction in the gap (Slavin & Madden,2001). Similarly, KIPP schools, a small network ofcharter schools that have extended hours, culturallyappropriate instructional practices, individualizedmentoring, and structured instruction, have shownpromising evidence of promoting racial-ethnicminority achievement (Educational Policy Institute,2005). These lines of evidence suggest that the qual-ity of instruction is a direct influence on theachievement gap, with differential allocation ofinstructional quality contributing to Black–Whitedifferences in learning and achievement.

In summary, several school-based resources havea critical influence on achievement, including thequality of teachers, curriculum, and teacher–studentrelationships. There is evidence that White students,on average, benefit from better teachers, more chal-lenging curriculum, and better relationships withteachers than their Black peers with similar recordsof achievement. Thus, these school factors likelyexert a direct influence on the Black–White achieve-ment gap.

Peers. Peer relationships and peer norms mayexert a direct influence on the achievement gap. Forexample, ethnographic research suggests that forBlack youth, the stigma associated with academicambition and its connection with White culture cancontribute to Black underachievement (Austen-Smith & Fryer, 2005; Fordham & Ogbu, 1986;Majors & Billson, 1993). Consistent with this formu-lation, Ferguson (2008) reported survey results sug-gesting that Black suburban high school students

were more likely than their White peers to reportwithholding academic effort because of concernabout how others might view them. Similarly,Graham (2001) reported that some racial-ethnicminority students value low-performing peers morethan high-performing peers. Other studies, how-ever, find less support for the effects of peer cultureon the achievement gap (Ainsworth-Darnell &Downey, 1998; Cook & Ludwig, 1998). In contrastto the literatures on the influence of family factorsand schooling, less empirical research has examinedthe influence of peers on the achievement gap. As aresult, it remains largely an open question whatpeer influences most strongly influence achievementat what age, whether those influences are presentin different amounts among children from differentracial-ethnic groups, and overall, how much peerinfluences contribute to the achievement gap.

Neighborhood. Neighborhoods may exert a directinfluence on the achievement gap. For example, the“collective efficacy” of neighborhoods, defined asthe level of community cohesion and social support,may influence a variety of outcomes, includingachievement (Sampson, Raudenbush, & Earls,1997). In communities with strong cohesion, neigh-bors support one another as they work to achieve arange of goals, including the education of their chil-dren. Cook, Herman, Phillips, and Settersten (2002)found that the higher the neighborhood cohesion,the better students performed in school. Further-more, children from different racial-ethnic groupsremain largely segregated from one another withinand across neighborhoods (Reardon et al., 2009)that differ in levels of achievement-promotingneighborhood cohesion. It is therefore plausible thatneighborhood cohesion would exert a direct influ-ence on the achievement gap. However, furtherresearch is necessary before a strong conclusion canbe drawn.

Factors outside the realm of direct influences. Manyphenomena partially fulfill the definition of a directinfluence, but are not direct influences as definedby SET. Many of these factors are critical contextsthat may themselves influence the proximal socialtransactions described by SET. For example, familystructure (McLanahan & Percheski, 2008), SES(Yeung & Conley, 2008), and family wealth (Orr,2003) are associated with achievement. In addition,there are mean racial-ethnic differences in familystructure, SES, and family wealth. These pheno-mena thus embody two characteristics of directinfluences: (a) they are associated with achieve-ment, and (b) they systematically differ betweenracial-ethnic groups.

1124 McKown

However, none of these phenomena is a socialprocess, defined as a communication between indi-viduals or between individuals and settings. As aresult, they do not meet SET’s definition of a directinfluence. SET is thus narrowly tailored to explainthe proximal social processes influencing theachievement gap. At the same time, there is verylittle doubt that broader contexts play a critical rolein shaping the achievement gap. SET conceptualizessocial processes as the mechanisms through whichbroader contexts shape the achievement gap. Thisconceptualization is consistent with work suggest-ing, for example, that the relation between SES anddevelopmental outcomes is mediated by more prox-imal social processes (Brooks-Gunn & Duncan,1997; Garcia-Coll, 1990; Guo, Brooks-Gunn, &Harris, 1996).

Summary. Social processes at home, at school, inthe peer group, and in neighborhoods are associ-ated with academic outcomes. Furthermore, thesesocial processes are distributed unequally to chil-dren from different racial-ethnic groups. SET pre-dicts that children from different racial-ethnicgroups with similar exposure to these direct influ-ences will achieve more similarly than childrenwho have different exposure to those direct influ-ences. However, SET also predicts that accountingfor multiple direct influences will, in general, reduce,but not eliminate racial-ethnic differences in achieve-ment. In addition to direct influences, other factors,called signal influences, contribute to the gap. As aresult, even a comprehensive account of directinfluences would not entirely explain the gap, par-ticularly in late elementary school and beyond.

Signal Influences on the Achievement Gap

Signal influences, defined. Signal influences aresocial events that signal to members of negativelystereotyped groups that they are devalued becauseof their group membership (Inzlicht & Ben-Zeev,2000; Mendoza-Denton, Downey, Purdie, Davis, &Pietrzak, 2002; Steele, 1997; Steele & Aronson,1995). Signal influences may be overt, such as whena person is told directly that people from his or hergroup are not capable. Signal influences may alsobe ambiguous, such as when one person communi-cates through nonverbal behavior negative feelingstoward another. Settings may also propagate signalinfluences outside of the context of an interpersonalencounter. For example, exposure to material thatmakes a child’s racial-ethnic group membershipsalient can trigger signal influences with no inter-personal interaction (Ambady, Shih, Kim, & Pittinsky,

2001). When the target of an interpersonal or settingsignal is from a stereotyped racial-ethnic group,the target may interpret that signal as a communi-cation that the target is devalued because of his orher racial-ethnic group membership. Such a signalcan interfere with performance in a stereotypeddomain (Steele & Aronson, 1995). To someonefrom a nonstereotyped group, the same socialevents will not be interpreted as devaluing a self-relevant identity.

Signal influences are transactional. They do notexist solely in the mind of stereotyped individuals.Neither do they exist solely in the social settingwith which the individual is engaged. Consistentwith the definition of stereotype threat (Steele &Aronson, 1995), signal influences occur when anevent in a social setting activates a concern in themind of a stereotyped individual—consciously ornot—that he or she may be judged on the basis ofthe stereotype. This process is distinct from self-stereotyping processes, which are presumed tooccur in the mind of the stereotyped individual. Itis also distinct from direct influences, althoughsome social processes, described later, can operateas either direct or signal influences.

A paradigmatic example from the stereotypethreat literature is that of the standardized testingsituation. Although outwardly the same for all chil-dren, when testing conditions highlight that a chal-lenging test is diagnostic of ability, members ofacademically stereotyped groups may become con-cerned that their performance will be judged inlight of a stereotype about their group’s intellectualability, in turn lowering performance (McKown &Strambler, 2009; McKown & Weinstein, 2003; Steele& Aronson, 1995). The same event may communi-cate different meanings to children, depending onthe prevailing cultural stereotype about each child’sracial-ethnic group. For children from academicallystereotyped racial-ethnic groups, the testing situa-tion may signal the possibility of being devalued,which, in turn, can hamper performance.

Routine signal events. What characteristics ofdevelopmental contexts are likely to propagate sig-nal influences on children’s achievement, producingracial-ethnic differences in achievement? Experi-mental evidence has identified several kinds ofevents that propagate signal influences, includingcharacterizing a test as diagnostic of natural ability(McKown & Strambler, 2009; McKown & Wein-stein, 2003, Study 2; Steele & Aronson, 1995, Experi-ment 1), cuing participants to think about astereotyped self-relevant identity before completinga task (Ambady et al., 2001), and telling partici-


pants directly that members of their group rou-tinely perform worse on a task than members ofother groups (Aronson et al., 1999). More common-place factors, such as group composition, can createsignal influences on academic performance, withminority status itself potentially signaling stereo-typed expectations to minority group members(Inzlicht & Ben-Zeev, 2000; Murphy, Steele, &Gross, 2007).

Signal influences have been consistently demon-strated in the lab. It is less clear what events indaily life communicate to children that their intel-lectual ability is devalued because of their racial-ethnic group. Research on interpersonal expectancyeffects suggests that teacher expectations may exertsignal influences. In many classrooms, teachersexpect more of White and Asian children thanBlack and Latino children with similar records ofachievement (Baron et al., 1985; McKown & Wein-stein, 2008; Rubovits & Maehr, 1973). Teacher dif-ferential treatment toward children from differentracial-ethnic groups can be interpreted by childrenas a signal that the intellectual ability of their groupis devalued. If this were so, the relation betweenteacher expectations and student achievementwould be stronger for children from stereotypedgroups than for their nonstereotyped peers. Indeed,the relation between teacher expectations and stu-dent achievement is more robust among Black thanWhite students (Jussim, Eccles, & Madon, 1996),with low teacher expectations associated morestrongly with negative academic outcomes for Blackthan White students (McKown & Weinstein, 2002).

It is conceivable that many events and inter-actions in the daily routine of schools signal tostudents from stereotyped groups that their intellec-tual ability is devalued. Those events may affect theextent to which students who are the targets ofstereotypes feel that school is fair. Indeed, correla-tional studies suggest that high levels of perceivedunfairness at school are related to higher rates ofdisciplinary referrals and disruptive behavior, andthis effect is particularly strong for Black and Latinostudents (Gottfredson, 2001; Gregory & Weinstein,2008; Tyler & Huo, 2000). When children from aca-demically stereotyped racial-ethnic groups believeteachers are unfair, this belief may activate culturalnarratives about racial injustice, signaling that theyare devalued because of their ethnicity. This may inturn have a negative impact on the academicachievement of children from stereotyped racial-ethnic groups.

Signal influences and interpretive skill. Signal influ-ences specifically and exclusively affect children

from negatively stereotyped groups. Negative stereo-types about the academic ability of Blacks persist(Bobo, 2001), although it is no longer sociallyacceptable to express those stereotypes directly(Dovidio & Gaertner, 1998). Furthermore, manypeople who hold overtly egalitarian beliefs nonethe-less may hold implicit stereotypes that lead them toappraise and respond to people on the basis of theirrace or ethnicity (Baron & Banaji, 2006; Gaertner &Dovidio, 1986; McConahay, Hardee, & Batts, 1981).Social norms against racism and the persistence ofimplicit stereotypes mean that in the contemporaryAmerican context, stereotypes are often expressedambiguously (Gaertner & Dovidio, 1986). For exam-ple, a White teacher who is formal and nervouswhen interacting with a particular Black studentmay be expressing a formal, socially anxious style.Alternatively, she may feel that Black students arenot capable, and her formality and discomfort areexpressions of this stereotypic belief (McKown &Strambler, 2009). How the student interprets theorigin of the teacher’s behavior is critical in deter-mining whether the teacher will propagate a signalinfluence.

The consequences of ambiguous signal eventsdepend on children’s awareness of stereotypesabout their group. For ambiguous signals to propa-gate signal influences, children must know thatothers might devalue their ability because of theirgroup membership. Equipped with this understand-ing, children can interpret ambiguous interpersonalbehavior as a reflection of racial-ethnic stereotypes.Although these capacities must be present forambiguous signals to have a negative influence onachievement, with adults and children, priorresearch suggests that stereotyped individuals donot have to be consciously aware of a signalinfluence for it to have a negative effect (Steele &Aronson, 1995). Rather, children and adults mustbe generally aware of self-relevant stereotypes inthe performance domain for ambiguous signals topropagate negative performance (McKown &Strambler, 2009; McKown & Weinstein, 2003).

Children’s thinking about others, about society,and about race and ethnicity changes dramaticallyin the first 12 years of life (Hirschfeld, 1996; Killen,Rutland, & Ruck, 2011; Killen & Stangor, 2001;Nesdale & Flesser, 2001; Quintana, 1998; Rutland,Cameron, Milne, & McGeorge, 2005). In turn, howchildren appraise and interpret social events affectstheir response to those events (e.g., McKown &Strambler, 2009; Spencer, 1999). By preschool,American children have developed a conception ofrace (Hirschfeld, 1996). By elementary school,

1126 McKown

children use group membership in their judgmentsabout the fairness of social exclusion (Killen, Lee-Kim, McGlothlin, & Stangor, 2002). In this agerange, children become aware of the relationbetween group membership and intergroup liking(Quintana, 1998). Specifically, between the ages of 6and 10, children become aware that others endorsestereotypic beliefs that may devalue their group’sintellectual ability (Brown & Bigler, 2005; Brown,Bigler, & Chu, 2010; McKown & Strambler, 2009;McKown & Weinstein, 2003; Quintana, 1998, 2008).

All these lines of evidence suggest that beginningaround third grade, and commonly by the end ofelementary school, typically developing childrenhave the social knowledge and capacity to interpretevents that, on their surface, are ambiguous interms of whether they reflect biased beliefs or atti-tudes, as expressions of stereotypes or prejudices.The ability to interpret events as signals about whatothers believe about a child’s group is, in turn,highly consequential.

For example, children’s growing stereotype con-sciousness can affect their response to social andacademic events. McKown and Strambler (2008)found that when children were aware of broadlyheld stereotypes, they were more likely to interpretnegative interethnic encounters as reflecting racialanimus. Furthermore, children from academicallynegatively stereotyped racial-ethnic groups (AfricanAmerican and Latino) who were aware of broadly heldcultural stereotypes performed more poorly on acognitive task when the task was described asdiagnostic of ability than when the same taskwas described as a problem-solving task. Forchildren from nonstereotyped racial-ethnic groupsand for children from stereotyped racial-ethnic groups who were unaware of broadly heldstereotypes, the conditions of testing did not affectperformance (McKown & Strambler, 2009; McKown& Weinstein, 2003). These findings suggest thatwhen children from stereotyped groups becomeaware of broadly held stereotypes, signal influencescan hamper performance on cognitive and aca-demic tasks.

Randomized field trials of interventions suggestthat children and youth can be inoculated againstnaturally occurring signal influences, further dem-onstrating their existence and impact. For example,drawing from on motivation theory (Dweck,1988), Good, Aronson, and Inzlicht (2003) foundthat teaching middle school girls to believe thatintelligence depends on effort and that academicdifficulties are situational reduced the effects ofstereotype threat on math test performance. Draw-

ing from self-affirmation theory (Steele, 1988),Cohen, Garcia, Apfel, and Master (2006) foundwhen African American students participated inan exercise in which they wrote a paragraphabout their values and why those values wereimportant, they showed lower levels of stereotypeactivation and received higher grades than theirAfrican American peers in a control condition. Bysuccessfully inoculating children against signalinfluences outside the lab, these studies indirectlydemonstrate the existence of signal influences innaturalistic settings.

Child Factors

The foregoing discussion of direct and signalinfluences emphasizes social processes and thesocial settings in which they take place. This raisesthe question of what role child factors play in caus-ing and maintaining racial-ethnic achievement gaps.There is little doubt that children’s individualbiological and cognitive characteristics affect theiracademic outcomes. For example, the better chil-dren perform on cognitive tasks such as those rep-resented in IQ tests, the better children doacademically (Neisser et al., 1995). Furthermore,how children understand and process information,and therefore what they can learn, changes norma-tively with age (Piaget, 1972).

It is beyond the scope of this article to reviewthe many ways individual child characteristics anddevelopmental processes contribute to achievement.An underlying premise of SET is that mean racial-ethnic group differences in achievement are theproduct of social processes that transpire betweenchildren and peers, between children and adults,and between children and social settings. Groupdifferences in IQ test scores are also presumed toarise from these same social forces. A corollary pre-mise of SET is that although individual child factorsinfluence within-group variability in achievement,they do not, in general, explain mean racial-ethnicgroup differences in achievement. There is oneimportant exception: The impact of signal influ-ences depends on children’s awareness of broadlyheld societal stereotypes, which changes predictablywith age (McKown & Strambler, 2008; McKown &Weinstein, 2003).

Overlap and Boundaries Between Direct and SignalInfluences

Direct and signal influences are distinct mecha-nisms by which a social event may affect group


differences on the Black–White achievement gap.Some events are clearly and only signal events. Forexample, the conditions of testing may signal tomembers of academically stereotyped racial-ethnicgroups that their intellectual ability is devalued.Those conditions create unequal outcomes exclu-sively through their signal properties. Other eventsare clearly and only direct influences. For example,enrollment in schools that deliver high-qualityinstruction is an important resource for the devel-opment of talent (Hamre & Pianta, 2005; Weinstein,2002). Higher quality instruction is associated withgreater opportunities to learn and achieve. To theextent that children from different racial-ethnicgroups are enrolled in different schools withinstruction whose quality is systematically different,this will exert an influence on racial-ethnic achieve-ment gaps through a direct influence on learningand achievement. If students are unaware that theallocation of resources is unequal, this form of dis-crimination will exert only a direct influence onracial-ethnic achievement gaps.

Some social phenomena can contribute to theachievement gap through direct influence, signalinfluence, or both (McKown, Gregory, & Weinstein,2010). Research on teacher expectations provides anillustration. Teachers who have higher expectationsof students tend to behave more warmly, providebetter instruction, and offer greater student auton-omy, all of which is associated with greater learn-ing (Brophy & Good, 1970; Weinstein, 2002). Insome classrooms, teachers expect more of theirWhite students than they do of their Black studentswith similar records of achievement (Baron et al.,1985; McKown & Weinstein, 2008). Teacher expecta-tions, allocated in part on the basis of child race,may lead teachers to provide better quality instruc-tion to White students than to Black students withsimilar records of achievement. This may in turnlead to Black–White differences in achievement. Inthis case, the differential allocation of instructionalquality is a direct influence, with no requirementthat children notice the differential treatment for itto contribute to the achievement gap. However, itis also possible that some students, noticing theteacher’s differential treatment of children from dif-ferent racial-ethnic groups, will interpret the tea-cher’s low expectations as reflecting the teacher’sbelief that some racial-ethnic groups are notcapable. If so, this may lead to a signal influence.Teacher expectations may thus exert direct and sig-nal influences on the Black–White achievement gap.

Although some experiences can operate as bothdirect and signal events, the distinction between the

two kinds of influences is nonetheless useful. Froma theoretical perspective, distinguishing direct andsignal influences provides a heuristic for evaluatingtwo distinct pathways through which events mayinfluence the achievement gap. From an appliedperspective, how we understand the mechanismsthrough which events affect the achievement gapwill shape the policies and practices that aredeployed to reduce the gap. For example, if teacherexpectations are seen as exclusively exerting adirect influence on the gap, then the best prescrip-tion for reducing its influence is to equalize teacherbehavior that communicates expectations. Attemptsto do so have met with mixed results (Gottfredson,Marciniak, Birdseye, & Gottfredson, 1995; Weinsteinet al., 1991). In contrast, if teacher expectations areseen as exerting a signal influence, then a prescrip-tion might be to inoculate students against theuntoward effects of signal influences by, for exam-ple, teaching children that achievement is the prod-uct of hard work (Good et al., 2003), or havingchildren affirm important positive self-relevant val-ues (Cohen et al., 2006). If, as the evidence sug-gests, teacher expectations can exert direct andsignal influences on the Black–White achievementgap, the focus might include equalizing teacherbehavior and inoculating students.

Combined Effects of Direct and Signal Influences

SET proposes that multiple social processes oper-ate together within and across key developmentalcontexts to influence racial-ethnic achievement gaps.Thus, with the goal of identifying the universe ofinfluences on the achievement gap, it is insufficientto account only for direct influences or only for sig-nal influences. It is also generally insufficient toaccount for direct and signal influences in a singlecontext. SET incorporates the proposition that directand signal influences can operate in any key devel-opmental context. The specific processes may besimilar or different in different contexts. In addition,the proportion of the variance in a group differenceresulting from direct and signal influences may bedifferent in each context and in each developmentalphase. However, the theory proposes that only byidentifying salient direct and signal influences inkey developmental contexts and by combiningeffects across contexts can we come to a fullaccounting of any racial-ethnic achievement gap.

Each of SET’s propositions is supported from theempirical literature. However, the overall proposi-tion that the achievement gap is better accountedfor by combining the effects of different social

1128 McKown

processes occurring in different key developmentalcontexts remains largely unexamined. Some research-ers have examined the combined association of morethan one direct influence on racial-ethnic achievementgaps (e.g., Brooks-Gunn et al., 1996, 2003; Lee &Burkam, 2002). However, no studies have examinedmultiple signal influences, or the combined effectsof direct and signal influences across contexts. SETlends itself to the study of multiple influencesacross multiple contexts simultaneously. It is con-ceivable that influences within and across develop-mental contexts would combine additively orinteractively to produce the Black–White achieve-ment gap. How varied forces described by SETcombine to produce the gap is an important areafor future investigation.

Universe of Influential Developmental Contexts

A key proposition of SET is that social processesthat cause and maintain the achievement gap mayoperate in and across multiple developmental con-texts (Bronfenbrenner, 1977; Kelly, 1987; Weinstein,2002). A context is defined as any setting in which anindividual transacts social interactions with impor-tant others or in which setting characteristics signalinformation of social value. Developmental contextsin childhood include family, school, peer network,and neighborhood. During different developmentalepochs, the contexts in which social processes influ-ence the achievement gap will change. As childrentransition to adulthood, new contexts become impor-tant, including the workplace and marriage. Withineach of these contexts are varied “microsettings,” inany one of which, direct or signal influences mayunfold. What contexts are the most important focusof inquiry and intervention should be guided by anunderstanding of the phenomenon of interest,theory, and the weight of evidence.

The earlier review of direct and signal influencessuggests that there is good reason to focus on fam-ily, school, peers, and neighborhoods. Specifically,the literature suggests that: (a) family factors suchas parenting practices influence the achievementgap; (b) some of the achievement gap can be attrib-uted to racial-ethnic differences in the availabilityof high-quality instruction; (c) signal influences innaturalistic settings, particularly school, contributeto the achievement gap; (d) peers may exert aninfluence on children’s academic achievement; and(e) neighborhood psychological resources influenceachievement and may influence the achievementgap. Furthermore, the foregoing literature reviewsuggests that direct and signal influences may

unfold in any one of these developmental settings.One of the most substantial unanswered questionsraised by SET is what direct and signal influencesoperate in what developmental settings at what ageto explain the achievement gap.

Developmental Hypotheses Uniquely DerivedFrom SET

SET leads to specific predictions about: (a) the rela-tive influence of direct versus signal influences onthe achievement gap at different ages, (b) whatdevelopmental contexts are likely to have the great-est influence on the achievement gap at what age,and (c) what set of direct and signal influences inwhat contexts are sufficient to account for theachievement gap.

Age-Related Differences in the Contribution of Directand Signal Influences

SET predicts that the relative impact of direct andsignal influences will change with age. In preschoolthrough second grade, most children are not yetaware that others endorse stereotypic beliefs (McKown& Strambler, 2009; McKown & Weinstein, 2003;Quintana, 1998, 2008). In those early years, then,only the most overt expression of stereotypes couldbe interpreted by a child as being self-relevant.Under these developmental constraints, childrencannot interpret more ambiguous cues that a childmay be judged on the basis of group membership asa devaluation of that group. By third grade, mostchildren have become aware of broadly held culturalstereotypes. Becoming aware of stereotypes is in turnassociated with susceptibility to stereotype threateffects induced through test instructions (McKown &Strambler, 2009; McKown & Weinstein, 2003).

SET thus predicts that in preschool through sec-ond grade, racial-ethnic differences in test scores areequal to the combination, across key developmentalcontexts, of direct influences only. Why? Signal influ-ences cannot affect children when they have not yetdeveloped the capacity to detect those signals. Aschildren develop the ability to interpret social cuesas reflecting stereotypes, a full explanation of theachievement gap requires accounting for both directand signal influences. Between third grade and ado-lescence, most children become aware of broadlyheld stereotypes. SET thus predicts that the relativeimpact of direct and signal influences on the Black–White achievement gap will shift from exclusivelydirect influences through second grade, to a prepon-


derance of direct influences and some signal influ-ences between third and fifth grades, to a more equalbalance between direct and signal influences thereaf-ter (see Figure 1).

The empirical literature is generally consistentwith this hypothesis. Among preschoolers, account-ing for multiple direct influences eliminates ornearly eliminates school readiness and test perfor-mance gaps (Brooks-Gunn et al., 2003; Lee &Burkam, 2002). In contrast, with middle school andhigh school students, controlling for direct influ-ences is insufficient to account entirely for theachievement gap (Duncan & Magnuson, 2005; butsee Mandara et al., 2009; Yeung & Pfeiffer, 2009).According to SET, this is because beyond secondgrade, controlling for direct influences on the Black–White achievement gap does not account for signalinfluences and thus leaves a residual portion of thegap unexplained. When children develop awarenessof broadly held cultural stereotypes, to account forthe entire gap requires accounting for both directand signal influences (McKown & Strambler, 2009;McKown & Weinstein, 2003).

Age-Related Differences in the Contribution of SocialInfluences Across Contexts

Prior work suggests what settings are most influ-ential in each epoch of a child’s life. In infancy andtoddlerhood, children spend most of their time andrelational energy with parents and other caregivers.After the transition to elementary school, first teach-ers, then peers, play increasingly prominent roles inchildren’s daily social experiences. Late in elemen-

tary school, as they become more independent, chil-dren may have more exposure to neighborhoodinfluences. In adolescence, close friendships, peernetworks, and romantic partners may becomeprominent influences, and parents exert less directinfluence over their children’s daily lives.

In light of these developmental changes, SEThypothesizes that prior to school entry, only homeinfluences affect the Black–White achievement gap.After school entry and through second grade,social processes at home and at school affect theBlack–White achievement gap, with home pro-cesses initially contributing more to the gap thanschool processes. As school becomes more academi-cally focused through fifth grade, social processesin school play an increasing role in contributing tothe gap. After second grade, peer relationshipsmove from more transient play to more enduring“chumships.” At this point, peers may begin tocontribute to the gap, and this influence willincrease in adolescence. In late elementary school,as children become more independent in the com-munity, neighborhood effects on the achievementgap come into play, and grow incrementallythrough adolescence. This hypothesis is depicted inFigure 2.

Intervention Implications

Developmental hypotheses derived from SETprovide the basis for creating a developmentallyinformed sequence of interventions. Specifically, toreduce the school readiness gap, preschool interven-tions should promote positive parenting andachievement-supporting home settings. From kin-dergarten through second grade, in addition to

Figure 1. Hypothesized relation between age and the proportionof the achievement gap accounted for by direct and signal influ-ences.

Figure 2. Hypothesized proportion of the achievement gap influ-enced in different developmental contexts, by age.

1130 McKown

family support, interventions should focus on pro-viding strong instruction and positive teacher–childrelationships. From third grade on, in addition toparent support and strong academic support, inter-ventions should inoculate children against theeffects of signal influences and should endeavor toreduce the overall level of stereotyping at school.From third grade on, interventions should also beimplemented to promote positive peer norms forachievement. From fifth grade on, in addition tothe aforementioned kinds of support, interventionsshould increase neighborhood collective efficacy or,if that is not feasible, help children cope with neigh-borhood influences. SET predicts that if direct andsignal influences salient in a particular setting and adevelopmental epoch are addressed, the Black–White achievement gap would be eliminated.

The Sufficiency of SET

Research on the achievement gap has beenmarked by a consistent partial accounting for thegap. SET holds promise to be a sufficient explanationfor the Black–White achievement gap across devel-opmental epochs. Not all possible developmentalprocesses and theories are incorporated into SET.Furthermore, other social forces not incorporatedinto SET may contribute to racial-ethnic achievementgaps. As a result, two forms of SET are possible. Ifdevelopmental phenomena and the full universe ofsocial forces are sufficiently accounted for by SET, astrong form of the theory will withstand empiricalscrutiny: Direct and signal influences, combinedacross key developmental contexts, will explain theentire Black–White achievement gap. If on the otherhand, relevant developmental processes or socialforces are not integrated into SET, a weak form ofthe theory may be more reasonable: Direct and sig-nal influences, combined across key developmentalcontexts, will explain more of the achievement gapthan current models do, but other forces, notaccounted for by SET also contribute to the gap.Whether the strong or weak form of SET is more jus-tified is a matter of empirical inquiry.

Standards of Evidence

Existing evidence supports many of the elements ofSET. However, further research is needed to evalu-ate the theory as a whole. A judicious combinationof naturalistic and experimental methods can beused to test the components of SET and the theoryin its entirety. Three standards are required, one for

each of the following: (a) evidence of direct influ-ences, (b) evidence of signal influences, and (c) evi-dence supporting the whole of SET.

Evidence of Direct Influences

The standard of evidence for the presence of adirect influence on the Black–White achievementgap is as follows: If a social process has a directinfluence, Black and White children who share simi-lar exposure to the direct influence should achievemore similarly than children with different expo-sure to the direct influence. Furthermore, the rela-tion between hypothesized direct influences andachievement should be similar for children from allracial-ethnic groups. For example, Black and Whitechildren who live in comparably cohesive neighbor-hoods should achieve more similarly than childrenwho live in neighborhoods that differ in their cohe-siveness. In statistical terms, controlling for directinfluences should reduce the magnitude of the rela-tion between racial-ethnic group membership andachievement, and there should be no ethnicity bydirect influence interaction.

Naturalistic research examining the relationbetween direct influences and the Black–Whiteachievement gap generally examines whether themagnitude of the relation between child race-ethnic-ity and achievement is reduced when hypothesizeddirect influences are included in a statistical model.For example, this method has been used to demon-strate that comparing children with similar homeenvironments accounts for much of the Black–Whiteachievement gap (Brooks-Gunn et al., 1996; Lee &Burkam, 2002). Further research using naturalisticmethods will help clarify the nature and magnitudeof the combined effect of direct influences on thegap. The benefit of naturalistic studies is that theytend to be more feasible than true experiments. Thedrawback is that they do not permit strong causalinferences.

True experiments can lead to inferences aboutthe causal impact of direct influences. For example,the New Hope Project, a randomized antipovertyprogram, found that income supplements, child-care assistance, and health-care benefits to full-timeworking poor families led to academic, social, andemotional benefits to children, particularly boys(Huston et al., 2001, 2005). Experiments permit stron-ger inferences about the causal relation betweenhypothesized direct influences and achievement.Although they are well suited to isolating the cau-sal influence of a single independent variable, trueexperiments can be costly.


Evidence of Signal Influences

If a setting characteristic has a signal influenceon the achievement gap, it should be true that inthe presence of the signal, members of stereotypedracial-ethnic groups only will perform more poorlythan when the signal is absent. In statistical terms,signal influences will produce an Ethnicity × SignalStrength interaction. More specifically, the strongerthe signal, the more poorly children from stereo-typed groups will perform; in contrast, childrenfrom nonstereotyped groups will perform similarlyregardless of the level of the signal.

As with direct influences, signal influences canbe assessed using experimental or naturalisticdesigns. With regard to experimental studies, a raftof studies on stereotype threat has demonstratedthat a variety of events can signal to children andadults from stereotyped groups that their intelli-gence is devalued, causing reduced performance(Nguyen & Ryan, 2008; Steele & Aronson, 1995;Walton & Spencer, 2009). Naturalistic methodscould also be used to examine whether members ofstereotyped groups, in the presence of presumedsignal events, are more negatively affected thanmembers of nonstereotyped groups. Experimentaland naturalistic methods have been used, for exam-ple, in studies demonstrating that teachers expectmore of White and Asian students than their Blackand Latino peers with similar records of achieve-ment (Baron et al., 1985; McKown & Weinstein,2008) and that the relation between teacher expecta-tions and achievement is stronger for Black stu-dents than for White students (Jussim et al., 1996),particularly when teacher expectations are negative(McKown & Weinstein, 2002).

Evidence Supporting SET as a Whole

Evidence supporting the constituent elements ofSET does not necessarily support SET as a whole.Many studies presented in this article support theconstituent elements of SET by demonstrating thata direct or signal influence accounts for part of theBlack–White achievement gap. However, no study hastotally accounted for the gap at all ages and nointervention or policy solution has totally elimi-nated it. A full assessment of SET will involveassessing the effect, across developmental contextsand ages, of multiple direct and signal influences.Ideally, such an examination would be undertakenusing naturalistic designs and field experiments.Doing so would provide a fair test of the theory asa whole and would provide useful information

about comprehensive strategies to reduce educa-tional inequality.

Conclusion

SET is a general, flexible model of social processes thatcontribute to the American racial-ethnic achievementgaps. SET proposes that racial-ethnic achievementgaps are the result of the combination, across keydevelopmental contexts, of direct and signal influ-ences. A central goal in developing SET is to pro-vide a parsimonious model from which hypothesescan be developed and systematically, programmati-cally, and empirically tested. Although the reviewof evidence presented in this article focused on theBlack–White achievement gap, SET provides aframework for understanding, examining, andreducing any racial-ethnic achievement gap. Fur-thermore, SET may prove useful in explainingsocioeconomic achievement gaps, and sex differ-ences in science, math, and engineering occupations(Cheryan & Plaut, 2010). SET may also offer a use-ful heuristic for understanding the origins ofinequalities in other domains, such as racial-ethnicdifferences in disciplinary referrals at school (Greg-ory, Skiba, & Noguera, 2010) and employment out-comes (Pager & Shepherd, 2008). The accuracy andgeneral applicability of SET to other forms ofinequality are empirical questions to be vigorouslyinvestigated.

Prior research on the Black–White achievementgap provides support for each of the constituentelements of SET. Skeptics will rightly concludethat support for SET’s constituent elements is notthe same as support for the theory as a whole.However, there have been no comprehensiveassessments of SET as a singular explanation of theBlack–White achievement gap. Although it posesmethodological challenges, only by conducting suchresearch can we understand the full range of socialprocesses that contribute to this significant socialproblem. In so doing, we may come to moreefficient and effective policies and practices toreduce inequalities that run afoul of core Americanideals of fair play.

References

Adler, N. E., Boyce, T., Chesney, M. A., & Cohen, S.(1994). Socioeconomic status and heath: The challengeof the gradient. American Psychologist, 49, 15–24.

Ainsworth-Darnell, J. W., & Downey, D. B. (1998).Assessing the oppositional culture explanation for

1132 McKown

racial/ethnic differences in school performance. Ameri-can Sociological Review, 63, 536–553.

Ambady, N., Shih, M., Kim, A., & Pittinsky, T. L. (2001).Stereotype susceptibility in children: Effects of identityactivation on quantitative performance. PsychologicalScience, 12, 385–390.

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 of Experimental Social Psychology, 35, 29–46.

Austen-Smith, D., & Fryer, R. G. (2005). An economicanalysis of “acting White.” Quarterly Journal of EconomicScience, 120, 551–583.

Baldwin, A. L., Baldwin, C., & Cole, R. E. (1990). Stress-resistant families and stress-resistant children. In J. E.Rolf, A. S. Masten, D. Cicchetti, K. H. Nuechterlein, &S. Weintraub (Eds.), Risk and protective factors in thedevelopment of psychopathology (pp. 257–280). New York:Cambridge University Press.

Baron, A. S., & Banaji, M. R. (2006). The development ofimplicit attitudes: Evidence of race evaluations from ages6 and 10 and adulthood. Psychological Science, 17, 53–58.

Baron, R. M., Tom, D. Y., & Cooper, H. M. (1985). Socialclass, race, and teacher expectations. In J. B. Dusek (Ed.),Teacher expectancies (pp. 251–270). Hillsdale, NJ: Erlbaum.

Baumrind, D. (1966). Effects of authoritative control onchild behavior. Child Development, 37, 887–907.

Baumrind, D., & Black, A. (1967). Socialization practicesassociated with dimensions of competence in preschoolboys and girls. Child Development, 38, 291–327.

Bobo, L. (2001). Racial attitudes and relations at the closeof the twentieth century. In N. J. Smelser, W. J. Wilson,& F. Mitchell (Eds.), America becoming: Racial trends andheir consequences (pp. 264–301). Washington, DC:National Academy Press.

Boykin, A. W. (1986). The triple quandary and the school-ing of Afro-American children. In U. Neisser (Ed.), Theschool achievement of minority children: New perspectives(pp. 57–92). Hillsdale, NJ: Erlbaum.

Brice-Heath, S. (1983). Ways with words: Language, life, andwork in communities and classrooms. New York:McGraw-Hill.

Bronfenbrenner, U. (1977). Toward an experimental ecol-ogy of human development. American Psychologist, 32,513–531.

Brooks-Gunn, J., & Duncan, G. (1997). The effects of pov-erty on children. The Future of Children, 7, 55–71.

Brooks-Gunn, J., Klebanov, P. K., & Duncan, G. J. (1996).Ethnic differences in children’s intelligence test scores:Role of economic deprivation, home environment, andmaternal characteristics. Child Development, 67, 396–408.

Brooks-Gunn, J., Klebanov, P. K., Smith, J., Duncan, G. J.,& Lee, K. (2003). The Black-White test score gap inyoung children: Contributions of test and family char-acteristics. Applied Developmental Science, 7, 239–252.

Brooks-Gunn, J., & Markman, L. B. (2005). The contribu-tion of parenting to ethnic and racial gaps in schoolreadiness. The Future of Children, 15, 139–168.

Brophy, J., & Good, T. (1970). Teachers’ communicationof differential expectations for children’s classroom per-formance: Some behavioral data. Journal of EducationalPsychology, 61, 365–374.

Brown, C. S., & Bigler, R. S. (2005). Children’s perceptionsof discrimination: A developmental model. Child Devel-opment, 76, 533–553.

Brown, C. S., Bigler, R. S., & Chu, H. (2010). An experi-mental study of the correlates and consequences ofperceiving oneself to be the target of gender discrimina-tion. Journal of Experimental Child Psychology, 107,100–117.

Cheryan, S., & Plaut, V. C. (2010). Explaining underrepre-sentation: A theory of precluded interest. Sex Roles, 63,475–488.

Clotfelter, C., Ladd, H., & Vigdor, J. (2004). Do schoolaccountability systems make it more difficult for lowperforming schools to attract and retain high qualityteachers? Journal of Policy Analysis and Management, 23,251–271.

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

Cook, T., Herman, M. R., Phillips, M., & Settersten, R. A.(2002). Some ways in which neighborhoods, nuclearfamilies, friendship groups, and schools jointly affectchanges in adolescent development. Child Development,73, 1283–1309.

Cook, P. J., & Ludwig, J. (1998). The burden of “actingWhite”: Do Black adolescents disparage academicachievement? In C. Jencks & M. Phillips (Eds.), TheBlack-White test score gap (pp. 375–400). Washington,DC: Brookings Institution.

Dovidio, J. F., & Gaertner, S. L. (1998). On the nature ofcontemporary prejudice: The causes, consequences, andchallenges of aversive racism. In J. L. Eberhardt & S. T.Fiske (Eds.), Confronting racism: The problem and theresponse. Thousand Oaks, CA: Sage.

Duncan, G. J., Brooks-Gunn, J., & Klebanov, P. K. (1994).Economic deprivation and early childhood develop-ment. Child Development, 65, 296–318.

Duncan, G. J., & Magnuson, K. A. (2005). Can familysocioeconomic resources account for racial and ethnictest score gaps? The Future of Children, 15, 35–54.

Dweck, C. S. (1988). Motivational processes affectinglearning. American Psychologist, 41, 1040–1048.

Educational Policy Institute. (2005). Focus on results: Anacademic impact analysis of the Knowledge Is Power Program(KIPP). Virginia Beach, VA: Author.

Entwisle, D. R., Alexander, K. L., & Olson, L. S. (1997).Children, schools, and inequality. Boulder, CO: Westview.

Farkas, F. (2003). Racial disparities and discrimination ineducation: What do we know, how do we know it, andwhat do we need to know? Teachers College Record, 6,1119–1146.

Ferguson, R. F. (2008). Toward excellence with equity: Anemerging vision for closing the achievement gap. Cam-bridge, MA: Harvard Education Press.


Fordham, S., & Ogbu, J. U. (1986). Black students’ schoolsuccess: Coping with the “burden of acting White.”Urban Review, 18, 176–206.

Fryer, R. G., & Levitt, S. D. (2004). Understanding theBlack-White test score gap in the first two years ofschool. Review of Economics and Statistics, 86, 447–464.

Fryer, R. G., & Levitt, S. D. (2005). The Black-White testscore gap through third grade (NBER Working Paper No.11049). Cambridge, MA: National Bureau of EconomicResearch.

Gaertner, S. L., & Dovidio, J. F. (1986). The aversive formof racism. In J. F. Dovidio & S. L. Gaertner (Eds.), Pre-judice, discrimination, and racism (pp. 61–89). San Diego,CA: Academic Press.

Garcia-Coll, C. (1990). Developmental outcome of minor-ity children: A process-oriented look into our begin-nings. Child Development, 61, 270–289.

Gonzales, N. A., Cauce, A. M., Friedman, R. J., & Mason,C. A. (1996). Family, peer, and neighborhood influencesamong African-American adolescents: One-year pro-spective effects. American Journal of Community Psychol-ogy, 24, 365–387.

Good, C., Aronson, J., & Inzlicht, M. (2003). Improvingadolescents’ standardized test performance: An inter-vention to reduce the effects of stereotype threat. Jour-nal of Applied Developmental Psychology, 24, 645–662.

Gottfredson, D. C. (2001). Schools and delinquency. NewYork: Cambridge University Press.

Gottfredson, D. C., Marciniak, E., Birdseye, A. T., & Gott-fredson, G. D. (1995). Increasing teacher expectationsfor student achievement. Journal of Educational Research,88, 155–163.

Graham, S. (2001). Inferences about responsibility and val-ues: Implication for academic motivation. In F. Salili &C. Y. Chiu (Eds.), Student motivation: The culture and con-text of learning. Plenum series on human exceptionality (pp.31–59). Dordrecht, the Netherlands: Kluwer Academic.

Gregory, A., Skiba, R. J., & Noguera, P. A. (2010). Theachievement gap and the discipline gap: Two sides ofthe same coin? Educational Researcher, 39, 59–68.

Gregory, A., & Weinstein, R. S. (2004). Connection andregulation at home and in school: Predicting growth inachievement for adolescents. Journal of AdolescentResearch, 19, 405–427.

Gregory, A., & Weinstein, R. S. (2008). The discipline gapand African Americans: Defiance or cooperation in thehigh school classroom. Journal of School Psychology, 46,455–475.

Guo, G., Brooks-Gunn, J., & Harris, K. M. (1996). Parents’labor force attachment and grade retention among urbanblack children. Sociology of Education, 69, 217–236.

Hamre, B. K., & Pianta, R. C. (2005). Can instructionaland emotional support in the first-grade classroommake a difference for children at risk of school failure?Child Development, 76, 949–967.

Hanushek, E. A., & Rivkin, S. G. (2009). Harming the best:How schools affect the black-white achievement gap.Journal of Policy Analysis and Management, 28, 366–393.

Hirschfeld, L. A. (1996). Race in the making: Cognition, cul-ture, and the child’s construction of human kinds. Cam-bridge, MA: MIT Press.

Hughes, J. N., Gleason, K. A., & Zhang, D. (2005). Rela-tionship influences on teachers’ perceptions of aca-demic competence in academically at-risk minority andmajority first grade students. Journal of School Psychol-ogy, 43, 303–320.

Humphreys, K., & Rappaport, J. (1993). From the commu-nity mental health movement to the war on drugs: Astudy in the definition of social problems. AmericanPsychologist, 48, 892–901.

Huston, A. C., Duncan, G. J., Granger, R., Bos, J.,McLoyd, V., Mistry, R., et al. (2001). Work-based antip-overty programs for parents can enhance the schoolperformance and social behavior of children. ChildDevelopment, 72, 318–336.

Huston, A. C., Duncan, G. J., McLoyd, V. C., Crosby, D.A., Ripke, M. N., Weisner, T. S., et al. (2005). Impactson children of a policy to promote employment andreduce poverty for low-income parents: New Hopeafter 5 years. Developmental Psychology, 41, 902–918.

Inzlicht, M., & Ben-Zeev, T. (2000). A threatening intellec-tual environment: Why females are susceptible to expe-riencing problem-solving deficits in the presence ofmales. Psychological Science, 11, 365–371.

Jencks, C., & Phillips, M. (Eds.) (1998). The black-white testscore gap. Washington, DC: Brookings Institution Press.

Jensen, A. (1969). How much can we boost IQ achieve-ment? Harvard Educational Review, 39, 1–123.

Jussim, L., Eccles, J. S., & Madon, S. (1996). Social percep-tion, social stereotypes, and teacher expectations: Accu-racy and the quest for the powerful self-fulfillingprophecy. Advances in Experimental Social Psychology, 28,281–388.

Kelly, J. G. (1987). An ecological paradigm: Defining men-tal health consultation as a preventive service. Preven-tion in Human Services, 4, 1–36.

Killen, M., Lee-Kim, J., McGlothlin, H., & Stangor, C.(2002). How children and adolescents evaluate genderand racial exclusion. Monographs of the Society forResearch in Child Development, 67(4), 1–119.

Killen, M., Rutland, A., & Ruck, M. (2011). Promotingequity, tolerance, and justice in childhood. SRCD SocialPolicy Report, 25, 3–25.

Killen, M., & Stangor, C. (2001). Children’s social rea-soning about inclusion and exclusion in gender andrace peer group contexts. Child Development, 82, 174–186.

Lee, J. (2002). Racial and ethnic achievement gap trends:Reversing the progress toward equity? EducationalResearcher, 31, 3–12.

Lee, V. E., & Burkam, D. T. (2002). Inequality at the start-ing gate: Social background differences in achievement aschildren begin school. Washington, DC: Economic PolicyInstitute.

Levin, H. M. (2009). The economic payoff to investing ineducational justice. Educational Researcher, 38, 5–20.

1134 McKown

Lucas, S. R., & Berends, M. (2002). Sociodemographicdiversity, correlated achievement, and de facto track-ing. Sociology of Education, 75, 328–348.

Magnuson, K., & Duncan, G. J. (2006). The role of familysocioeconomic resources in the Black-White test score gapamong young children. Developmental Review, 26, 365–399.

Majors, R., & Billson, J. M. (1993). Cool pose: The dilemmasof Black manhood in America. New York: TouchstoneBooks/Simon & Schuster.

Mandara, J., & Murray, C. B. (2002). Development of anempirical typology of African American family func-tioning. Journal of Family Psychology, 16, 318–337.

Mandara, J., Varner, F., Greene, N., & Richman, S. (2009).Intergenerational family predictors of the Black-Whiteachievement gap. Journal of Educational Psychology, 101,867–878.

McConahay, J. B., Hardee, B. B., & Batts, V. (1981). Hasracism declined in America? It depends on who is ask-ing and what is asked. Journal of Conflict Resolution, 25,563–579.

McKown, C. (2005). Applying ecological theory toadvance the science and practice of school-based preju-dice reduction interventions. Educational Psychologist,40, 177–189.

McKown, C., Gregory, A., & Weinstein, R. S. (2010).Expectations, stereotypes, and self-fulfilling propheciesin classroom and school life. In J. Meece & J. Eccles(Eds.), Handbook of research on schools, schooling, andhuman development. New York: Routledge.

McKown, C., & Strambler, M. J. (2008). Social influenceson the ethnic achievement gap. In S. Quintana & C.McKown (Eds.), Handbook of race, racism, and the develop-ing child (pp. 366–396). Hoboken, NJ: Wiley.

McKown, C., & Strambler, M. J. (2009). Developmentalantecedents and social and academic consequences ofstereotype-consciousness in middle childhood. ChildDevelopment, 80, 1643–1659.

McKown, C., & Weinstein, R. S. (2002). Modeling the roleof child ethnicity and gender in children’s differentialresponse to teacher expectations. Journal of AppliedSocial Psychology, 32, 159–184.

McKown, C., & Weinstein, R. S. (2003). The developmentand consequences of stereotype-consciousness in mid-dle childhood. Child Development, 74, 498–515.

McKown, C., & Weinstein, R. S. (2008). Teacher expecta-tions, classroom context, and the achievement gap.Journal of School Psychology, 46, 235–261.

McLanahan, S., & Percheski, C. (2008). Family structureand the reproduction of inequality. Annual Review ofSociology, 34, 257–276.

Mendoza-Denton, R., Downey, G., Purdie, V. J., Davis,A., & Pietrzak, J. (2002). Sensitivity to status-basedrejection: Implications for African American students’college experience. Journal of Personality and Social Psy-chology, 83, 896–918.

Murphy, M. C., Steele, C. M., & Gross, J. J. (2007). Signal-ing threat: How situational cues affect women in math,science, and engineering settings. Psychological Science,18, 879–885.

Neisser, U., Boodoo, G., Bouchard, T. K., Boykin, A.W.,Brody, N., Ceci, S.J. (1995). Intelligence: Knowns andunknowns. American Psychologist, 51, 77–101.

Nesdale, D., & Flesser, D. (2001). Social identify and thedevelopment of children’s group attitudes. Child Devel-opment, 72, 506–517.

Nguyen, H. D., & Ryan, A. M. (2008). Does stereotypethreat affect test performance of minorities andwomen? A meta-analysis of experimental evidence.Journal of Applied Psychology, 93, 1314–1334.

Oakes, J. (2005). Keeping track: How schools structure inequal-ity (2nd ed.). New Haven, CT: Yale University Press.

Ogbu, J. (2002). Cultural amplifiers of intelligence: IQ andminority status in cross-cultural perspective. In J. M.Fish (Ed.), Race and intelligence: Separating science frommyth (pp. 241–278). Mahwah, NJ: Erlbaum.

Orr, A. J. (2003). Black-White differences in achievement: Theimportance of wealth. Sociology of Education, 76, 281–304.

Page, L. C., Murnane, R. J., & Willett, J. B. (2008). Trendsin the Black-White achievement gap: Clarifying the meaningof within and between-school achievement gaps (NBERWorking Paper No. 14213). Cambridge, MA: NationalBureau of Economic Research.

Pager, D., & Shepherd, H. (2008). The sociology of dis-crimination: Racial discrimination in employment,housing, credit, and consumer markets. Annual Reviewof Sociology, 34, 181–209.

Phillips, M., Brooks-Gunn, J., Duncan, G. J., Klebanov, P.,& Crane, J. (1998). Family background, parenting prac-tices, and the Black-White test score gap. In C. Jencks& M. Phillips (Eds.), The Black-White test score gap (pp.103–145). Washington, DC: Brookings Institution Press.

Phillips, M., Crouse, J., & Ralph, J. (1998). Does the Black-White test score gap widen after children enter school?In C. Jencks & M. Phillips (Eds.), The Black-White testscore gap (pp. 229–272). Washington, DC: BrookingsInstitution Press.

Piaget, J. (1972). The psychology of intelligence. Totowa, NJ:Littlefield Adams.

Quintana, S. M. (1998). Children’s developmental under-standing of ethnicity and race. Applied and PreventivePsychology, 7, 27–45.

Quintana, S. M. (2008). Racial perspective taking ability:Developmental, theoretical, and empirical trends. In S.M. Quintana & C. McKown (Eds.), Handbook of race, rac-ism, and the developing child. Hoboken, NJ: Wiley.

Reardon, S. F., Farrell, C. R., Matthews, S. A., O’Sulliv-and, D., Bischoff, K., & Firebaugh, G. (2009). Race andspace in the 1990s: Changes in the geographic scale ofracial residential segregation, 1990–2000. Social ScienceResearch, 38, 55–70.

Reardon, S. F., & Galindo, C. (2009). The Hispanic-Whiteachievement gap in math and reading in the elemen-tary grades. American Educational Research Journal, 46,853–891.

Reardon, S. F., & Robinson, J. P. (2007). Patterns and trendsin racial/ethnic and socioeconomic academic achieve-ment gaps. In H. F. Ladd (Ed.), Handbook of research ineducation finance and policy. New York: Routledge.


Rubovits, P. C., & Maehr, M. L. (1973). Pygmalion inBlack and White. Journal of Personality and Social Psy-chology, 25, 210–218.

Rushton, J. P., & Jensen, A. R. (2005). Thirty years ofresearch on race differences in cognitive ability. Psychol-ogy, Public Policy, and Law, 11, 235–294.

Rutland, A., Cameron, L., Milne, A., & McGeorge, P.(2005). Social norms and self-presentation: Children’simplicit and explicit intergroup attitudes. Child Develop-ment, 76, 451–466.

Sackett, P. R., Hardison, C. M., & Cullen, M. J. (2004). Oninterpreting stereotype threat as accounting for AfricanAmerican-White differences on cognitive tests. AmericanPsychologist, 59, 7–13.

Saft, E. W., & Pianta, R. C. (2001). Teachers’ perceptionsof their relationships with students: Effects of child age,gender, and ethnicity of teachers and children. SchoolPsychology Quarterly, 16, 125–141.

Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997).Neighborhoods and violent crime: A multilevel studyof collective efficacy. Science, 227, 918–924.

Seidman, E. (1983). Unexamined premises of social prob-lem solving. In E. Seidman (Ed.), Handbook of socialintervention (pp. 48–71). Beverly Hills, CA: Sage.

Slavin, R. E., & Madden, N. A. (2001). Reducing the gap:Success for all and the achievement of African American stu-dents Presented at the annual meeting of the AmericanEducational Research Association. Seattle, WA.

Spencer, M. B. (1999). Social and cultural influences onschool adjustment: The application of an identity-focused cultural ecological perspective. Educational Psy-chologist, 34, 43–57.

Steele, C. M. (1988). The psychology of self-affirmation:Sustaining the integrity of the self. In L. Berkowitz(Ed.), Advances in experimental social psychology (Vol. 20,pp. 261–302). San Diego, CA: Academic Press.

Steele, C. M. (1997). A threat in the air: How stereotypesshape intellectual identity and performance. AmericanPsychologist, 52, 613–629.

Steele, C. M., & Aronson, J. (1995). Stereotype threat andthe intellectual test performance of African Americans.Journal of Personality and Social Psychology, 69, 797–811.

Tharp, R. G. (1989). Psychocultural variables and con-stants: Effects on teaching and learning in school. Amer-ican Psychologist, 44, 349–359.

Thernstrom, A., & Thernstrom, S. (2002). No excuses: Closingthe racial gap in learning. New York: Simon & Schuster.

Tyler, T. R., & Huo, Y. J. (2000). How different ethnicgroups react to legal authority. San Francisco: Public Pol-icy Institute of California.

Vanneman, A., Hamilton, L., Anderson, J. B., & Rahman,T. (2009). Achievement gaps: How Black and White stu-dents in public schools perform in mathematics and readingon the national assessment of educational progress (NCESNo. 2009-455). Washington, DC: U.S. Department ofEducation.

Walton, G. M., & Spencer, S. J. (2009). Latent ability:Grades and test scores systematically underestimate theintellectual ability of negatively stereotyped students.Psychological Science, 20, 1132–1139.

Weinstein, R. S. (2002). Reaching higher: The power of expec-tations in schooling. Cambridge, MA: Harvard Univer-sity Press.

Weinstein, R. S., & McKown, C. (1998). Expectancy effectsin context: Listening to the voices of students and teach-ers. In J. Brophy (Ed.), Advances in research on teaching(Vol. 7, pp. 215–242). Greenwich, CT: JAI Press.

Weinstein, R. S., Soulé, C. R., Collins, F., Cone, J., Mehl-horn, M., & Simontacchi, K. (1991). Expectations andhigh school change: Teacher-researcher collaboration toprevent school failure. American Journal of CommunityPsychology, 19, 333–364.

Yeung, W. J., & Conley, D. (2008). Black-White achieve-ment gap and family wealth. Child Development, 79, 303–324.

Yeung, W. J., & Pfeiffer, K. M. (2009). The Black-Whitetest score gap and early home environment. Social Sci-ence Research, 38, 412–437.

1136 McKown

social equity theory and racial-ethnic achievement gaps€¦ · social equity theory and...

Documents