assessing the opositional culture explanation for racial differences in school

8/13/2019 Assessing the Opositional Culture Explanation for Racial Differences in School

http://slidepdf.com/reader/full/assessing-the-opositional-culture-explanation-for-racial-differences-in-school 1/19

Assessing the Oppositional Culture Explanation for Racial/Ethnic Differences in SchoolPerformanceAuthor(s): James W. Ainsworth-Darnell and Douglas B. DowneySource: American Sociological Review, Vol. 63, No. 4 (Aug., 1998), pp. 536-553Published by: American Sociological AssociationStable URL: http://www.jstor.org/stable/2657266

Accessed: 17/02/2009 09:43

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at

http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

http://www.jstor.org/action/showPublisher?publisherCode=asa.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the

scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that

promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

American Sociological Association is collaborating with JSTOR to digitize, preserve and extend access to

American Sociological Review.

http://www.jstor.org

http://www.jstor.org/stable/2657266?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/action/showPublisher?publisherCode=asa

http://www.jstor.org/action/showPublisher?publisherCode=asa

http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/stable/2657266?origin=JSTOR-pdf



ASSESSING THE OPPOSITIONAL CULTURE EXPLANATION FOR

RACIAL/ETHNIC DIFFERENCES IN SCHOOL PERFORMANCE*

JamesW.Ainsworth-Darnell DouglasB. DowneyThe Ohio State University The Ohio State University

The oppositional culture explanation for racial disparities in school perfor-

mance posits that individuals from historically oppressed groups (involun-

tary minorities) signify their antagonism toward the dominant group by re-

sisting school goals. In contrast, individuals from the dominant group and

groups that migratedfreely to the host country (immigrant minorities) main-

tain optimistic views of their chances for educational and occupational suc-

cess. Because of its historical and cross-cultural appeal, this explanation

has been well-received by academics, although key implications of the theory

have not been carefully tested. Proponents have failed to systematically com-

pare perceptions of occupational opportunity and resistance to school across

involuntary, dominant, and immigrant groups. Using a large sample of Afri-

can American, Asian American, and non-Hispanic white high school sopho-

mores from the first follow-up of the National Education Longitudinal Study,

we provide the first rigorous test of the oppositional culture explanation.

Upon close scrutiny, its key predictions fail.

R espite recent improvement on somemeasures, the gap in educationalper-

formance across racialgroupspersists. Find-ing explanations for that gap continues tofrustrateacademics. Some scholars point tocharacteristicsof the minority family itself(Moynihan 1965), while others see differ-ences in educational performanceas prima-rily a function of social structuralconditions(Bourdieu 1977; Bowles and Gintis 1976),

such as the types of neighborhoods studentslive in (Massey and Denton 1993) and con-sequently the kinds of schools they attend.The oppositional culture explanation drawsfrom both of these traditions, recognizingthat social structuralconditions shape oppor-tunities but arguing that these conditionsform students' motivation for schooling.

*Direct all correspondence to Douglas B.

Downey, Department of Sociology, 300 Bricker

Hall, 190 N. Oval Mall, Columbus, Ohio 43210([email protected]). This research was sup-ported by a Spencer Postdoctoral Fellowship toDowney. The authors' contributions are equal.We appreciatethe comments of Lowell Hargens,Robert Kaufman, Lauren Krivo, RoslynMickelson, Brian Powell, Rob Robinson,Catherine Ross, Ruben Rumbaut, and MaureenTobin.

However,we challenge the maintenets of theoppositional cultureexplanation.

THE OPPOSITIONAL CULTUREEXPLANATION

Ogbu's (1978, 1991a) explanation for racialdifferences in school performance, referredto here as the oppositional culture explana-tion (or the resistance model), has gained

considerable acceptance among scholars. Akey component in the explanationis the dis-tinction between immigrant minorities-

groups who migratedto the host country oftheir own free will-and involuntary minori-

ties-groups historically enslaved, colo-nized, or conquered. Immigrant minoritiestend to compare their condition to that ofrelatives in theirhomelands,and because thiscomparison is usually favorable, they de-velop optimistic attitudes regarding both

their chances for success in the new countryand the payoff for efforts aimed at promot-ing achievement. In comparison,involuntaryminorities are in a psychologically vulner-able position; their membersdid not migratewith an expectation to improve their condi-tion, but were incorporated into societyagainst their will. Lacking an identifiable

536 American Sociological Review, 1998, Vol. 63 (August:536-553)



RACIAL/ETHNIC DIFFERENCES IN SCHOOL PERFORMANCE 537

foreign reference group, involuntaryminori-ties contrast their condition with that of thedominant group. This comparison tends toproduce resentment because they conclude

that they fare poorly solely because they be-long to a disfavored group. Involuntarymi-norities learn from those around them thatthey have limited job opportunities, and sothey put forth little effort toward success inschool because, as proponents of this expla-nation contend, there is a reciprocal relation-ship between the opportunities available to aminority group and the patternof linguis-tic, cognitive, motivational, and other

school-related skills they develop (Ogbu1978:5; also see Ogbu 1991a for a moreelaborate description of the oppositionalcul-ture explanation).

Sociologists find attractive several compo-nents of Ogbu's (1978; 1991a) explanationfor racial disparities in school performance.The central thesis, that perceptions of occu-pational opportunityshapestudents'personalcharacteristics, such as their motivationallevels and value for schooling, provides acompelling link between societal conditionsand actors' daily actions. Of course, a simi-lar argumenthas been made by others to ex-plain the attitudesandbehaviorsof working-class students. For example, the lads inWillis's (1976) Learning to Labor rejectedschoolwork because it was perceived as hav-ing little consequence for their future lives.Similarly, Kohn (1977) notes that working-class parents emphasize characteristics in

their childrenthatwill serve them well in theroles they expect their children to fill asadults-largely working-class positions.Both Willis and Kohn stress that the verycharacteristics working-class students de-velop also prevent them from excelling inschool and moving into middle-class occu-pations. Ogbu applies this same logic to in-voluntary minorities, contendingthatpercep-tions of poor occupational opportunitiesen-

courageresistance to school

goals.Perhaps because of its sociological ap-peal, the resistance model has been ap-plauded by a wide range of scholars(Erickson 1987; Fischer et al. 1996; Foley1991; Jaynes and Williams 1989), eventhough the model's key claims have not re-ceived empirical verification. Specifically,proponents have failed to systematically

compare perceptions of occupational oppor-tunity and resistance to school across invol-untary, dominant, and immigrantgroups. Toaddress this problem, we assess whether

four of the model's key claims, derived pri-marily from ethnographic studies, are con-sistent with data from a national survey.

Our data come from the first follow-up ofthe National Education Longitudinal Study(NELS), a national sample of almost 17,000high school sophomores collected in 1990 bythe National Center for Education Statistics(NCES). We focus on the 2,197 AfricanAmerican, 653 Asian American, and 13,942

non-Hispanic white students in our samplebecause these groups are the best representa-tives of involuntary, immigrant, and domi-nant groups in the United States.' Althoughour examination of four hypotheses derivedfrom the oppositional culture model is shortof an exhaustive test, lack of supportfor anyof these claims is cause for skepticism.

THE OPPOSITIONALCULTUREMODEL:

FOUR HYPOTHESESAND RESULTS

Hypothesis 1: Involuntary minority (AfricanAmerican) students perceive fewer returnsto education and more limited occupa-tional opportunities than do dominant(white) students or immigrant minority(Asian American) students.

Ogbu (1978) writes, An importantdeter-minant of school performanceis what chil-

dren and their parents or community expectto gain from their education in adult life (p.54). He cites his own work and that of otherethnographers as evidence that AfricanAmericansperceive limited occupationalop-

1 The Asian American group includes Chinese,Filipino, Japanese, and Korean students. Weoriginally kept these Asian subgroups separate,but because they behaved similarly in our modelswe combined them. We excluded SoutheastAsians because their status as an immigrant mi-nority is less clear. A small numberof immigrantblacks was excluded.

In practice, distinguishing between involuntaryand immigrant minorities is not as straightfor-ward as oppositional culture proponents wouldlike (e.g., see Tuan 1995), but because our goal isto assess oppositional culture claims, we employthe proponents' categorizations.



538 AMERICAN SOCIOLOGICAL REVIEW

Table 1. Means, Standard Deviations, and Descriptions for Dependent Variables Used in the Analysis:

High School Sophomores from National Education Longitudinal Study, 1990

Variable name Description Metric Mean S.D. Alpha

Educational Outcome

Student grades Self-reported grades in math, 0 = Mostly below D 4.63 1.49 .77English, history, and science in all academicclasses. subjects taken;

7 = Mostly A's in all

academic subjectstaken.

Perceptions of Future Opportunity

Education is Do you agree with the following 0 = Strongly disagree; 2.57 .61

important to statement about why you go 3 = Strongly agree.getting a job to school? Education is important

later on for getting a job later on.

Occupational Occupation that you expect to 0 = Not planning to 2.43 1.29expectations have at age 30. work, don't know;

4 = Professional/clergy.

Skills, Habits, and Styles

Effort Standardized scale of teachers' -8.63= Lowest stan- -.13 2.29 .88responses to: Does this student dardized score;

usually work hard for good grades? 3.47 = Highest stan-

(Yes/No); How often does this dardized score.

student complete homework

assignments? (Never-all of the

time); How often is this studentattentive in class? (Never-all ofthe time).

Disruptive Teachers' responses to: How often 0 = Never; .67 .76is this student disruptive in class? 4 = All of the time.

In trouble How many times did the following 0 = Never to all five 1.75 3.07 .64

things happen to you in the first questions;half of the current school year? 50 = Ten times or more

(a) I got in trouble for not follow- to all five questions.ing school rules; (b) I was put onan in-school suspension; (c) I was

suspended or put on probation fromschool; (d) I was transferred toanother school for disciplinaryreasons; (e) I was arrested.

Homework Overall about how much time do 0 = None; 4.25 4.14you spend on homework each 16 = Over 15 hours.week out of school?

Concrete Attitudes

Treatment by Do you agree that: (a) When you 0 = Disagree strongly 5.38 1.52 .64teachers work hard on schoolwork, your with a and c, and

teachers praise your efforts; (b) In agree strongly with b;

class you often feel put down by 9 = Agree stronglyyour teachers; (c) Most of your with a and c, and

teachers really listen to what you disagree stronglyhave to say. with b.

Attitude toward Do you agree that: (a) The teaching 0 = Disagree strongly 3.75 1.08 .54a

teachers is good at your school; (b) Teachers with both statements;are interested in students. 6 = Agree strongly with

both statements.

(Table I continued on next page)




(Table I continued from previous page)

Variablename Description Metric Mean S.D. Alpha

Concrete Attitudes (Continued)

Discipline is fair How much do you agreewith the 0 = Stronglydisagree; 1.68 .69following statementabout your 3 = Stronglyagree.current chool and teachers?Discipline is fair.

OKto breakrules How often do you thinkit is OKto: 0 = Never to all four 2.41 2.26 .78(a) cut a couple of classes; (b) skip questions;school for a whole day; (c) talk 12 = Often to all fourback to teachers; d) disobey school questions.rules?

Doing whatI am Do you agreewith the following 0 = Stronglydisagree; 1.83 .65supposed o do statementsaboutwhy you go to 3 = Stronglyagree.

in class school?I get a feeling of satisfac-tion fromdoing whatI'm supposedto do in class.

OK to cheat How often do you think it is OK to: 0 = Never to both 1.48 1.42 .63a(a) cheat on tests; (b) copy some- questions;one else's homework? 6 = Often to both

questions.

Good student Do you think that other students 0 = Not at all; 1.18 .60see you as a good student? 2 = Verymuch.

Troublemaker Do you think that other students 0 = Not at all; .33 .55see you as a troublemaker? 2 = Verymuch.

Tries hard n class How often do you tryas hardas 0 = Never in all 3.15 .94 .75you can in math, English, history, academicsubjectsand science? taken;

4 = Almost every day inall academicsubjectstaken.

Popularity among Peers

Popularity Do you think that other students 0 = Not at all to all 2.88 1.36 .74see you (a) as popular; b) as threequestions;socially active; (c) as partof the 6 = Verymuch to allleadingcrowd? three questions.

Note: Valid cases rangefrom 11,937 to 16,972.a Bivariatecorrelation.

portunities (Dollard 1957; Ogbu 1974:97;Powdermaker1968). We question these stud-ies, however, because they lack a direct com-parison with white students and becausesome surveys indicate more favorable occu-pational expectations among African Ameri-cans (Porter 1974; Portes and Wilson 1976).In addition, the extensive literature docu-menting African Americans' optimistic edu-cational expectations (Solorzano 1991)raises the possibility that African Americansmay be similarly optimistic regarding theiroccupational chances.

To test this issue more explicitly, we mea-sure whether youths reportthat education is

important or getting a job later on, and wegauge perceptions of occupational opportu-nity by asking about the type of occupationthe youth expects to have at age 30. We re-gress these two dependent variables on racein unadjustedmodels (Model 1), and then inmodels controllingfor background Model 2)and socioeconomic characteristics (Model3). Table 1 presents detailed information onall dependentvariables.

Because we are interested in racial/ethnicdifferences in performancethat are indepen-dent of socioeconomic factors, we include inour models factors other than immigrant/in-voluntary minority status that may affect




Table 2. Unstandardized Coefficients from the Regression of Perceptions of Future Opportunity onRace and Selected Independent Variables: High School Sophomores from the National Edu-cation Longitudinal Study, 1990

Education s Importantfor Gettinga Job LaterOn OccupationalExpectations

IndependentVariables Model 1 Model 2 Model 3 Model 1 Model 2 Model 3

African American .042*** .025* .042*** .049* .081** .150***(.012) (.013) (.013) (.024) (.026) (.026)

Asian American .104*** .093** .086** .343*** .274*** .250**(.030) (.030) (.030) (.063) (.063) (.062)

Family income (in $10,000s) .004*** .023***

(.001) (.003)

Parental occupational prestige .028*** .065***

(.006) (.013)Parental education .028*** .1 17***

(.004) (.009)

R2 .001 .015 .025 .002 .026 .057

Note: Numbers n parentheses re standard rrors.White is the omittedcategoryfor race. Models2 and 3also include the following control variables:sex, region, numberof siblings, parents' age, family structure,and urban/suburban/ruralocation of school.

* <.05 **p< .01 p < .001 (two-tailed tests)

educational performance.We control for par-ents' highest level of education (O = no highschool to 6 = advanceddegree), parents' oc-cupational prestige (O = low occupationalstatus to 4 = high occupational status),2andfamily income (in tens of thousands of dol-lars). In addition, we include controls forsex, region, numberof siblings, parental age,family structure, and urban/suburban/rurallocation of school because these variablesmay be associated with race and may predict

educational performance. In each analysis,white students are the omitted category, andfor missing data we substituted the mean.3We also weighted the data to account for theNELS sampling design.4

Our results for Hypothesis 1, presented inTable 2, contradict the oppositional culturemodel's claim: In the adjusted models(Model 3), African American students aresignificantly more likely than white studentsto report that education is important to get-ting a job later on (b = .042) and to havemore optimistic occupational expectations(b = .150) than white students. Therefore, akey assumption of the oppositional culturemodel-that involuntaryminorities perceive

lower returns to education and more limitedoccupational opportunities than do studentsfrom the dominant group-is not supported.

Hypothesis 2: Involuntary minority (AfricanAmerican) students exhibit greater resis-tance to school than do dominant (white)students or immigrant minority (AsianAmerican) students.

Consistent with Hypothesis 2, many eth-

nographers have described conflict between

2 For parents' education, occupation, and age,we used informationfrom both parents wheneverboth were available. The coded data for the vari-ables measuring parents' education and parents'occupational prestige reflect the highest levels ofeducation and prestige achieved by both parents.

Parents' age is coded as the average age of thetwo parents.

3 Binary variables indicating missing-variablestatus (Cohen and Cohen 1983) were not statisti-cally significant. We also estimated our modelsusing listwise deletion of missing data and foundsimilar patterns.

I Most statistical packages compute standard

errors that are technically too small when analyz-ing data based on a clustered sampling designsuch as that used for the NELS (NCES 1995). Wepresent standard errors unadjusted for researchdesign, which provide the most favorable test ofthe oppositional culture hypotheses. The overallpatterns are identical, however, whether or not wecorrected the standarderrors.




involuntary minority students and teachersand administrators Gilmore 1985; Solomon1992; Weis 1985). While these ethnographiesoffer rich descriptions of the behaviors of

some groups of students, critics of the oppo-sitional culture model may remain un-convinced because these studies lack a care-ful comparisonof African Americanstudentsandwhite students'resistance toward school.It is possible that many high school stu-dents-not just African Americans-sharesubstantialanti-academicvalues, as Colemanreported (1961).

Measuring resistance to school poses a

special challenge. Ogbu (1991b) contendsthat attitudinal indicators of resistance toschool may not be adequate:

[O]ne learns what blacks believe about howthey get ahead in America not necessarily byasking them direct questions about gettingahead; direct questions will generally elicit re-sponses similar to those given by white Ameri-cans. A more useful approach is to observewhat they do in orderto get ahead. (P. 444)

Others agree that observing what studentsdo-their skills, habits, and styles -is keyto understanding school success (Swidler1986), but students' attitudes and preferencesare not inconsequential because they signal

competence and count for good citizen-ship (Lamont and Lareau 1988). Mickelson(1990) notes that measures of students' atti-tudes regarding specific, everyday events-what she calls concrete attitudes-can pre-dict educational performance. To gauge re-

sistance to school, we consider students'skills, habits, and styles and their concreteattitudes.

We measure skills, habits, and styles withindicators similar to those used by others(Farkas 1996). Two teachers assessed eachstudent's classroom effort and whether thestudentwas disruptive in the classroom. Wealso use two student-reportedmeasures: (1)In trouble is a composite variable con-structedfrom several questions about eventsthat happened to the youth during the firsthalf of the 1989-1990 school year (e.g., gotin troublefor not following school rules;wassuspended);and(2) homeworkrepresentsthenumberof hours the studentspenton school-work per week.

We measure students' concrete attitudeson a broad range of specific issues regard-

ing school. Our attitudinal variables coverstudents' perceptions of how teachers treatthem (treatment by teachers) and their moregeneral attitude toward teachers. We also

tap into potential frustrations with beingmistreated by asking students whether theythink discipline is fair at their school. Andbecause involuntary minorities are expectedto be especially resistant to school rules, wecreated three variables: OK to break rules,whether the student reports a feeling of sat-isfaction from doing what I am supposed todo in class, and OK to cheat on tests andcopy homework. We also use the items Do

you think that other students see you as agood student? (good student), and alterna-tively, Do you think that other students seeyou as a troublemaker? troublemaker).Fi-nally, tries hard in class gauges students'self-reported effort. To determine whetherAfrican American students resist schoolgoals more than do white studentsandAsianAmerican students, we regress our indica-tors of skills, habits, and styles and concreteattitudes on race in unadjusted models(Model 1) and in models controlling forbackground (Model 2) and socioeconomicvariables (Model 3).

We find mixed results when testing Hy-pothesis 2. When we focus on students'skills, habits, and styles, our results are con-sistent with the oppositional culture model:From Model 3 in Table 3a, African Ameri-can students are evaluated by their teachersas putting forth significantly less effort (b =

-.332) and as more frequently being disrup-tive (b = .129) than white students.And stu-dents' self-reports agree-African Ameri-cans report doing less homework (b =-.274) than do white students. AfricanAmerican students also report being introuble more often in the unadjustedmodel(b = .185), but this patternis accounted forby the background variables included inModel 2.

For our measures of concrete attitudes,however, the results are strikingly different.Overall, African American students reportmore positive attitudes toward school thando white students (Table 3b). For example,focusing on the adjusted model (Model 3),African American students are significantlymore likely than their white counterpartstoreport good treatment by teachers (b =




Table 3a. Unstandardized Coefficients from the Regression of Skills, Habits, and Styles on Race andSelected Independent Variables: High School Sophomores from the National Education Lon-gitudinal Study, 1990

Effort DisruptiveIndependentVariable Model 1 Model 2 Model 3 Model 1 Model 2 Model 3

AfricanAmerican -.563*** -.444*** -.332*** .154*** .140*** .129***(.048) (.052) (.051) (.013) (.014) (.015)

Asian American .393** .293* .235 -.1 15*** -.1 10 -.101* '(.124) (.124) (.122) (.035) (.035) (.035)

Family income .018*** .001(in $10,000s) (.005) (.002)

Parental occupational .086*** .011prestige (.025) (.007)

Parental education .270*** -.048**(.018) (.005)

R .009 .059 .085 .009 .056 .062

In Trouble Homework

IndependentVariable Model 1 Model 2 Model 3 Model 1 Model2 Model 3

AfricanAmerican .185** .050 -.016 -.526*** -.465*** -.274**(.065) (.070) (.070) (.081) (.088) (.087)

Asian American -.755*** -.536** -.499** 1.345*** .716 ** .632*(.167) (.167) (.167) (.210) (.210) (.207)

Family income -.008 .047*-*

(in $10,000s) (.007) (.009)

Parental occupational -.024 .205***

prestige (.035) (.043)

Parental education -.184*** 370- -

(.025) (.031)

R2 .002 .041 .046 .005 .043 .066

Note: Numbers n parenthesesare standard rrors.White is the omittedcategoryfor race. Models 2 and 3also include thefollowing control variables:sex, region,numberof siblings, parents'age, family structure,and urban/suburban/ruralocation of school.

*p < .05 **p< .01 ***p .001 (two-tailed tests)

.289), less likely to agree that it is OK tobreak rules (b = -.450), and more likely toreport a feeling of satisfaction from doingwhat they are supposed to do in class (b =.129). Similarly, African Americans are sig-nificantly less likely than whites to agreethat it is OK to cheat (b = -.236) and more

likely to report that others view them as agood student (b = .125) but not a trouble-maker (b = -.070). Finally, African Ameri-can students are significantly more likelythan their white counterpartsto report thatthey try hard in class (b = .120). Overall,African American students and AsianAmerican students consistently report morepro-school attitudes than do white students.

Results for discipline is fair provide theonly exception.5

Hypothesis 3: High-achieving involuntaryminority (African American) students arenegatively sanctioned by their peers fortheir achievement.

Oppositional culture proponents recog-nize that not all involuntary minority stu-dents do poorly in school. Proponents claim,

I A reviewer suggested that resistanceto schoolmay be distributed more unevenly among Afri-can Americans than for whites, with high resis-tance among the most disadvantaged and strongschool support among the middle and upper so-




Table 3b. Unstandardized Coefficients from the Regression of Concrete Attitudes on Race and Se-lected Independent Variables: High School Sophomores from the National Education Lon-gitudinal Study, 1990

Treatment by Teachers Attitude toward teachers Discipline Is FairIndependent Variable Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3

African American .327*** .264*** .289*** .061** .018 .044 -.060*** -.075*** -.062'***

(.032) (.035) (.035) (.023) (.025) (.025) (.013) (.015) (.015)

Asian American .194* .108 .106 .116* .058 .054 .100** .066 .061(.083) (.084) (.084) (.059) (.060) (.060) (.035) (.035) (.035)

Family income .016*** .014*** .004**

(in $10,000s) (.004) (.003) (.002)

Parental occupational .026 .037** -.003

prestige (.018) (.012) (.007)

Parental education .016 .018* .030***(.012) (.009) (.005)

R2 .006 .014 .016 .001 .007 .012 .002 .007 .011

Doing What I AmOK to Break Rules Supposed to Do In Class OK to Cheat

IndependentVariable Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3

African Americans -.502*** -.428*** _.450:?** .135*** .119*** .129*** -.302*** -.237*** -.236***

(.048) (.052) (.052) (.012) (.014) (.014) (.030) (.033) (.033)

Asian American -.169 -.209 -.195 .110*** .087** .087** -.079 -.068 -.066(.123) (.124) (.124) (.032) (.032) (.032) (.077) (.078) (.078)

Family income - -.001 .006*** .005

(in $10,000s) (.006) (.00 1) (.003)

Parental occupational - -.040 .014* -.047**

prestige (.026) (.007) (.016)

Parental education - -.050** .004 .015

(.018) (.005) (.012)

R2 .007 .035 .037 .007 .013 .015 .006 .020 .020

Good Student Troublemaker Tries Hardin Class

IndependentVariable Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3

African American .092*** .105*** .125*** -.084*** -.067*** -.070*** .145*** .123*** .120***

(.011) (.012) (.012) (.010) (.011) (.011) (.018) (.020) (.020)

Asian American .194*** .159*** .150 -.071 ** -.065* -.061 * .018 -.000 .002

(.029) (.030) (.029) (.027) (.027) (.027) (.047) (.047) (.047)

Family income .004** .002 - -.000

(in $10,000s) (.001) (.001) (.002)

Parental occupational .027*** -.001 .011

prestige (.006) (.006) (.010)Parental education .039*** - -.020*** - -.014*

(.004) (.004) (.007)

R2 .006 .021 .035 .004 .045 .047 .004 .030 .030

Note: Numbers in parentheses are standard errors. White is the omitted category for race. Models 2 and 3

also include the following control variables: sex, region, number of siblings, parents' age, family structure,and urban/suburban/rural location of school.

*P<.05 ** < .01 ***p< .001 (two-tailed tests)




Table 4. Unstandardized Coefficients from the Regression of Popularity on Race and Selected Inde-

pendent Variables: High School Sophomores from the National Education Longitudinal

Study, 1990

Popularity

Independent Variable Model 1 Model 2 Model 3 Model 4 Model 5

African American .010 .042 .079* .024 -.027(.029) (.032) (.032) (.031) (.036)

Asian American -.177* -.164* -.170* -.235** -.167(.074) (.076) (.076) (.075) (.093)

Family income (in $1,000s) .020*** .018*** .018***

(.003) (.003) (.003)

Parental occupational prestige .094*** .083*** .083***(.016) (.016) (.016)

Parental education .006 -.010 -.009

(.011) (.011) (.011)

Does youth think that other students see him/her as a good student?

Very good -373*** .340***(.026) (.029)

Not at all good -.542*** -.534***(.039) (.040)

African American .227***x Very good student (.067)

Asian American -.154

x Very good student (.156)

African American -.089x Not at all good student (.137)

Asian American -.368x Not at all good student (.446)

R2 .000 .004 .011 .040 .040

Note: Numbers in parentheses are standard errors. White is the omitted category for race; somewhat

good is the omitted category for good student. Models 2 through 5 also include the following controlvariables: sex, region, number of siblings, parents' age, family structure, and urban/suburban/rural location

of school.<.05 *p < .01 < .001 (two-tailed tests)

however, that those involuntary minoritystudents who do excel tend to feel psycho-logical pressure from the burden of actingwhite. Fordham and Ogbu (1986) explainthat because African Americans develop an

oppositional social identity that maintainsboundaries between themselves and whites,

they regard certain activities or events,symbols, and meanings as not appropriate

for them because those behaviors, events,symbols, and meanings are characteristicofwhite Americans (p. 181). Because per-forming well in school is defined as

white, high-achieving African Americansfeel burdened because other African Ameri-cans perceive them as selling out and

acting white, a pattern also reportedamong Chicano Mexican-oriented students(Portes and Zhou 1993:89).

While ethnographers find indications ofpeer sanctioning, we question whether thispattern prevails in a representative sample.To test whether the general cultural normagainst academic achievement in high schoolis any more prevalentamong African Ameri-

cioeconomic groups. We find no evidence of thisin the NELS data. Instead, our results moreclosely follow the patternreportedby Hochschild(1995), showing less variance across socioeco-nomic levels in the African American samplethan in the white sample.




cans thanit is among whites or Asian Ameri-cans, we predict students' popularity bywhether they are labeled a good student byothers. Our composite measure of student's

popularity is based on three questions: Doyou thinkotherstudents see you as (1) popu-lar, (2) socially active, and (3) part of theleading crowd?

Table 4 presents the results of predictingpopularitywith good student, race, and in-teractions between good student and race.For all students, those who reportthat othersview them as a very good versus a some-what good (omitted category) or not at all

good student also report the most popular-ity among their peers. But with respect to theoppositional culture model, our tests contra-dict Fordhamand Ogbu's (1986) claims. Wefind a positive interactionbetween being Af-ricanAmerican and being viewed as a verygood student(b = .227), demonstrating hat,relative to white students,African Americanstudentsare especially popularwhen they arealso seen as very good students.6 Further-more, in supplemental analyses (not shown),we comparedstudents' responses to the ques-tion, Among the friends you hang out with,how important is it to: study, get goodgrades,finish high school, and continue theireducationbeyond high school. Again, Afri-can American students report more pro-school values among their peers than dowhite students, and this discrepancy in-creases in models controlling for backgroundand socioeconomic differences.

Hypothesis 4: Resistance to school accountsfor the racial gap in school performancebetween involuntary (African American)students and dominant (white) studentsand immigrant minority (Asian American)students.

Although their ethnographic evidence isconsistent with this claim, Ogbu and associ-ates (Ogbu 1978; Fordham and Ogbu 1986)have not systematically tested whether re-sistance to school explains the racial gap inperformance. We predict students' overall

6 Supplemental analyses revealed that this pat-tern is robust, whether high-achieving studentsare defined by their standardizedtest scores, orby grades, or by using self-concept as the depen-dent variable.

grades for math, English, history, and sci-ence classes in unadjustedmodels. Then weintroduce our measures of resistance toschool indicators of skills, habits, and

styles and concrete attitudes as blocks ofintervening variables, both separately andtogether.

Table 5 shows the results of this analysis.Note that at the bivariate level (Model 1),African Americans have poorer grades (b =-.254) than do white students, and AsianAmericans have better grades (b = .530).Controlling for background and socioeco-nomic differences (Model 3) reduces these

disparitiessubstantially.And when we intro-duce indicators of skills, habits, and styles(Model 4) the African American coefficientis reduced to nonsignificance, supportingthe emphasis others have placed on the roleof students' culturalskills for explaining ra-cial differences in school outcomes (Farkas1995). Skills, habits, and styles are also im-portant for explaining Asian Americans'success the Asian American coefficientdrops from .364 to .241 on their inclusion.

Model 5, which predicts grades from con-crete attitudes (omitting skills, habits, andstyles), exacerbates the gap between AfricanAmericans and whites (b = -.087 in Model3, b = -.280 in Model 5), largely becauseAfrican Americans generally have morepro-school attitudes. These attitudinal indi-cators are also key to understanding he suc-cess of Asian American students: The coef-ficient for Asian Americans is reduced by

roughly one-half, from .364 (Model 3) to.171 (Model 5). Clearly, it would be a mis-take to conclude that racial differences inschool performance are solely a result ofvariations in school behavior. Without stu-dents' concrete attitudes in the model, wewould err by assuming that African Ameri-cans would earn similar grades if their be-haviors were like white students'. This isnot true. In Model 4, African American stu-dents are statistically equal to whites withrespect to grades, but their attitudes, whichare not controlled, are much more positive.Yet when African American students haveattitudes only as positive as whites (Model6), they earn lower grades. Without suchpro-school attitudes, African American stu-dents would surely be farther behind whitestudentsthanthey are now.




Table 5. Unstandardized Coefficients from the Regression of Student Grades on Race and Selected

Independent Variables: High School Sophomores from the National Education Longitudi-

nal Study, 1990

Student Grades

Independent Variable Model 1 Model 2 Model 3 Model 4 Model 5a Model 6a

African American -.254*** -. 1744** -.087** .017 -.280*** -. 130***(.029) (.032) (.031) (.026) (.027) (.025)

Asian American .530*** .404**:* .364* .241*** .171 .149**(.075) (.075) (.074) (.062) (.063) (.058)

Family income (in $10,000s) .019*** .01 I*** .013*** .009***(.003) (.003) (.003) (.003)

Parental occupational prestige -I.09 1** .056*1* .057*** .042***(.015) (.013) (.013) (.012)

Parental ducation .1-79* .076*** .132*** .074***

(.01 1) (.009) (.009) (.009)

Skills, Habits,and Styles

Effort .2754z** .214***(.004) (.004)

In trouble -.054*** -.034***

(.003) (.003)

Homework .049*** .028***

(.002) (.002)

ConcreteAttitudesTreatment by teachers .035*** .020**

(.007) (.007)

Attitude toward teachers .037*** .014(.010) (.010)

Discipline is fair .011 -.015

(.015) (.014)

OK to break rules -.025*** .010*

(.005) (.005)

Doing what I am supposed .144*** .127***

to in class (.017) (.016)

OK to cheat .003 -.010

(.007) (.007)

Good student 1.053* * .774***(.018) (.017)

Troublemaker -. 178*** -.028(.020) (.019)

Tries hard in class .072*** .008

(.01 1) (.010)

R2 .008 .041 .080 .355 .328 .442

Note: Numbers in parentheses are standard errors. White is the omitted category for race. Models 2

through 6 also include the following control variables: sex, region, number of siblings, parents' age, family

structure, and urban/suburban/rural location of school. Disruptive was omitted from this analysis due to

colinearity issues.

a When we include perceptions of future opportunity in Models 5 and 6 as additional measures of atti-

tudes toward school, the African American coefficient is -.292 in Model 5 and - .144 in Model 6.

*p < .05**p < .01 ***p< .001 (two-tailed tests)




SUMMARY EVALUATION OF THEOPPOSITIONAL CULTUREEXPLANATION

Our goal has been to evaluate key claims ofthe oppositional culture explanation for ra-cial differences in school performance. Us-ing data from a national survey, we foundthat the model is inconsistent with the datain several ways. The fundamental flaw ofOgbu's (1978; 1991a) oppositional cultureexplanation is that African American stu-dents do not perceive fewer returnsto edu-cation and more limited occupational oppor-

tunities than do whites. Without support forthis cornerstone of the theory, AfricanAmericans' relative lack of skills, habits, andstyles is open to alternative explanations.Also, while others have shown that AfricanAmericans reportrelatively high educationalexpectations (Solzarno 1991), we demon-strate that this patternextends much further.On a host of specific questions about theireveryday lives in the classroom, AfricanAmerican students report more pro-schoolattitudes than do white students. And ratherthan suffering sanctioning from peers, Afri-can Americans who are viewed as good stu-dents are more likely to be popularthan aretheir white counterparts.

How do we reconcile our results with thosegarneredfrom the many ethnographies thatdescribe anti-school attitudesamongAfricanAmericans (Gilmore 1985; Ogbu 1978;Solomon 1992; Weis 1985)? One way is to

restrict our sample to the most discouragedstudents: dropouts. Fortunately, the NELSlocated and surveyed many of these drop-outs, so we were able to compare dropouts'occupational expectations and attitudes to-ward school with those of the rest of thesample. Focusing only on those AfricanAmericans who dropped out of school be-tween the eighth and tenth grades reveals agroup of African Americans much like theones Ogbu describes frustrated with theiroccupationalchances, pessimistic about theirfutures, and resistant to school goals (resultsavailable from authorson request).

Because the goal of the oppositional cul-ture model is to explain societal racial dif-ferences in school performance(Ogbu 1978,1991a), we contend that tests of the fullrange of African American students rather

thanthose focusing only on the most discour-aged are more appropriate.We agree that,under some conditions, African Americansmay see little profit in continuing their edu-

cations, in partbecause they perceive limitedopportunities in the labor market. Yet whenwe analyze a representative group of AfricanAmericans, we see patterns that contradictthe oppositional culture model. It is impor-tant, therefore, not to misconstrue the prob-lems of the most disadvantaged AfricanAmericans as necessarily characteristic ofthe experiences of all AfricanAmericans.

The second reason why our results are in-

consistent with Ogbu and his colleagues isthat we systematically compare AfricanAmerican students with their white counter-parts. This careful comparison of attitudestoward school with a representative samplereveals thatAfrican American students havemore pro-school attitudes than do whites.Again, this type of comparisonis not incom-patiblewith ethnographicresearch.MacLeod(1995) observed the attitudes and behaviorsof two groups of teenagers living in a low-income neighborhood the mostly white

Hallway Hangers and the mostly blackBrothers. He compared their perceptions

of occupational opportunity and foundgreater optimism among the African Ameri-can group. Interestingly, comparing occupa-tional expectations and attitudes towardschool among African American dropoutswith those of white dropouts in our sampleproduces patterns similar to those in Tables

2, 3a, and 3b, although the differences areless likely to reach statistical significance be-cause of small samplesizes. Even among thisdisenfranchised group, therefore, AfricanAmericans maintain more pro-school atti-tudes than do whites (results available fromthe authorson request).

RETHINKING THE AFRICANAMERICAN ATTITUDE-

ACHIEVEMENT PARADOX

How can we make sense of African Ameri-cans' optimistic occupational expectationsand positive concrete attitudes yet relativelypoor classroom behavior as indicated byskills, habits, and styles? African Americansappearto value school but fail to put forththe necessary effort for success. We assess




three possible explanations: (1) abstract ver-sus concrete attitudes, (2) positivity bias, and(3) values versus material conditions.

Abstract versus Concrete Attitudes

Mickelson's (1990) distinction between atti-tudes that matter for educational perfor-mance and those that do not could explainour results. She explains that researchershave been puzzled by African American stu-dents' pro-education attitudes yet relativelypoor school achievement because they mea-sure attitudes with abstract indicators (e.g.,

Education is the key to success in the fu-ture )that simply reflect the dominant ideol-ogy rather hanconcrete indicators rooted nlife experience (Mickelson 1990:51).Mickelson claims to resolve this paradoxby noting that African American studentssupport education at the abstract level butfeel frustratedwith schooling at the concretelevel, and that concrete attitudes predict stu-dents' grades, while abstract attitudes donot.7 Similarly, Ogbu (1991b) explains awayblacks' pro-school attitudes by claiming thatthey represent wishful thinking and thatblacks simplydo not match theiraspirationswith effort (p. 446). As a result, althoughethnographic researchers and survey re-searchers agree that blacks report pro-schoolattitudes, Ogbu (1991b) dismisses these atti-tudes as unimportant.

But our data, and that of many others,demonstrate that stated preferences, or atti-

tudes, do matter for success in school. Re-searchers have consistently documentedAf-rican Americans' strong educational aspira-tions, but to be consistent with Mickelson'sandOgbu's explanations, educationalaspira-tions would have to have little impact oneducationalperformance.This is not the case(Kerckhoff and Campbell 1977). Similarly,in our study African Americans reportedmore pro-school attitudes for nearly all ofour attitudinal

measures,and these same in-

dicators were meaningful predictors of edu-cational success above andbeyond the effectof behaviors.

I Note that we borrowed Mickelson's concretetermto highlight the way in which our attitudinalindicators measure students' views of their owneveryday school experiences.

In addition, we find that correlations be-tween attitudes and behaviors for blacks areonly slightly lower than those for whites. Insupplementalanalyses using the NELS data,

we tested whetherstudents' occupational ex-pectations influence skills, habits, and stylesand concrete attitudes differently for AfricanAmerican students than for white students.The interactionswere not significant and pro-duced no consistent pattern. We also testedwhether African Americans' pro-school atti-tudes have a smaller effect on their gradesthan do pro-school attitudes for grades ofwhite students.This claim was supportedfor

only two of our nine indicators of school atti-tudes ( good student, troublemaker ).Insum, compared to white students, AfricanAmerican students' responses to our ques-tions regarding attitudes toward school cor-related equally well with occupational expec-tations and nearly as well with grades.8 Theparadoxcannot be resolved, therefore, by ar-guing that the pro-school attitudes reportedby African Americans are meaningless foreducational success.

Finally, by simply noting thata given atti-tude is rarely the only independent variablepredictinga behavior,we can understandhowblacks can have morepositive attitudes(validones) andpoorerbehaviors (valid ones) thanwhites. In a multivariatemodel, if blacks aredisadvantagedon otherindependentvariablesthat predict behaviors, (e.g., materialcondi-tions and other unmeasured variables) thenthey could easily have a higher group mean

on attitudes but a lower group mean on be-haviors. There is no need to invoke an atti-tude-behavior inconsistency argument.

Positivity Bias

An alternativeexplanation is that black stu-dents are, on average, reluctant to presentthemselves negatively to a research enter-prise conductedby whites. Although within-race attitude-behavior correlations

maybe

similar, the argumentis that African Ameri-cans' responses to both attitudinal and be-

8 For many attitude-behavior relationships,Asian American correlations more closely matchthose of African Americans than those of whites.Based on predictive validity, therefore, AsianAmericans' responses are no more credible thanthose of African Americans.




havioral survey questions reflect positivitybias and are therefore less trustworthy.Forexample, blacks may report working veryhard n school but, on average, they may not

put forth the same effort as white studentswho report the same attitude.

Actual as opposed to correlational atti-tude-behavior inconsistency is not easily de-termined, however. For example, in theNELS data African American students whoreported that they work as hard as they canalmost every day in their classes also re-ported doing an average of 3.9 hours ofhomework per week. In contrast, Asian

American students and white students whoreportedworking as hard as they can almostevery day averaged 7.5 and 5.4 hours ofhomework per week, respectively. It is un-clear whether black students' attitudes areinconsistent with their behaviors, however,because there is no definite numberof hoursstudents who work as hard as they can al-most every day are supposed to spend onhomework.9 Consequently, we are no morejustified in concluding that blacks exhibitpositivity bias than that whites suffer fromnegativity bias.

We emphasize this point because scholarsconsidering this issue tend to start with theassumption that white Americans' attitude-behavior relationship is normative and thatdepartures from this norm are deviant.10Consider an alternative scenario that as-sumes that African American's attitude-be-havior relationship is normative.The para-

dox then becomes, Given their relativelystrong achievement, why do white students

have such anti-school attitudes? Our pointis that for the types of attitudinal questionswe use in our study, determining whetherblacks' reported attitudes are more positive

than they should be or whether white's re-ported attitudes are more negative than theyshould be is not a resolvable question.

Of course, black students and white stu-dents may employ different reference groupswhen answering attitudinal questions.Whether one tries hard in school maymean different things to black students andwhite students because they are surroundedby different sets of peers with varying norms

and expectations regarding school-relatedbehaviors. Recognizing that the meaning ofsurvey questions may differ for black stu-dents and white students, however, does notjustify labeling African Americans' re-sponses as less credible than those of whitesor Asian Americans.

Values versus Material Conditions

Continuedoccupationaldiscrimination(Far-ley 1984), residential segregation (Masseyand Denton 1993), anddifferentialtreatment(Cose 1993; Feagin 1991) should foster frus-tration and resentment among AfricanAmericans rather than optimism. In manyways, African Americans do express contin-ued frustrationwith currentracial conditions.For example, Hochschild (1995) explainsthat AfricanAmericans

...are more sure than whites that racial dis-

crimination inhibits black Americans ...

more pessimistic about how much successblacks can anticipate . . ., more convinced thatblacks' life chances are not within their con-trol . . . , and slightly less confident that theycontrol their own life chances.... Neverthe-less, African Americans remain more confidentthan whites about their own prospects. (P. 69)

So while cognizant of and frustrated bypersisting inequalities, African Americanstend to maintain optimism regarding theirown personal chances for success. Why?First, even disenfranchised groups tend toshare the values of the dominant ideology(Hochschild 1995; Wilson 1996), and so it isnot surprising that African Americans arealso shaped by the rhetoric of the AmericanDream. Second, African Americans as a

9 This pattern may also simply reflect a biasamong teachers in assigning more homework towhite students and Asian American students.

10This assumption may stem from some sur-veys that suggest that blacks' responses are lessreliable than whites. For example, Jones andForrest (1992) report that blacks are more likelythan whites to underreport abortions. Similarly,Fendrich and Vaughn (1994) and Alexander,Entwisle, and Bedinger (1994) note that blacksless accurately recall prior information. Whilethese studies suggest that blacks' responses tosurveys may, in some cases, be less reliable thanthose of whites, it is not clear whether this pat-tern extends to our analyses. Of course, even ifblacks' reports in the NELS are less reliable thanthose of whites, that is still no justification forcategorically dismissing them.




grouphistoricallyhave placed greatvalue oneducation. Indeed, the struggle for equal ac-cess to education is a central element of Af-ricanAmericangroupidentity. Because there

have been some tangible victories (Brown v.Board of Education,and the Civil Rights Actof 1964), African Americans maintain opti-mism that they personallywill be able to suc-ceed, although they hesitate to expect suc-cess for fellow blacks. Third, our data dem-onstrate that pro-school attitudes matter forsuccess. The rationalstudent,therefore,maystrategicallyadoptpro-school values becausethis is one element of the educationalprocess

within their control.On the surface, our data suggest thatAfri-

can Americans believe in the AmericanDreambut do not act like they do. To under-standhow a group may have relatively posi-tive school attitudes yet relatively poorschool behaviors, it is useful to think abouthow a group's material condition mightshape its members' tool kit of culturalskills (Swidler 1986). AfricanAmerican stu-dents tend to live in neighborhoods withmaterial conditions (e.g., high unemploy-ment and nontraditional family structures)that are less likely to foster the kinds ofskills, habits, and styles that lead to schoolsuccess. For example, contrast the life ofstudent A, who lives in a world in whichparents rise daily to prepare for work andother adults in the neighborhoodalso followthe structuredroutine of work, with studentB, whose parents and many other adults in

the neighborhood are unemployed. Otherfactors being equal, studentA will be morelikely than studentB to develop the habit ofbeing on time to school because of exposureto and emphasis on a daily schedule. De-spite valuing education, therefore, AfricanAmerican students are less likely to exhibitthe kinds of school-related behaviors thatteachers reward because [s]kills, habits,and styles are often shaped by the frequency

at which they are found in their own com-munity (Wilson 1996:72).1l

1 Although some Asian groups have developed

skills, habits, and styles useful for success in

school despite limited material conditions, theyhave not experienced the levels of economic dep-rivation or residential segregation that African

Americans have endured.

DISCUSSION

The oppositional culture model has becomeso respected in the academic community that

it threatens to divert attentionfrom other ex-planations for the racial gap in school per-formance. We hope our results, which con-tradict key claims, will stimulate a new dia-logue. Understanding racial differences inschool performance remains an importantgoal, but generatinga credible answer to thisissue was not our goal-our more modestaim was to assess the merits of one specificexplanation. Still, our study has important

implications for scholars studying racial andethnic differences in school performance.We agree with Ogbu (1978, 1991b) that

materialconditionsshape students'behaviorsin school, primarilyby influencing the kindsof skills, habits, and styles they develop. Andwe note that these behaviors are strong pre-dictors of educational performanceand sub-stantially mediate racial/ethnic effects ongrades (Farkas 1995). Distinct from Ogbu,however, we contend that students' prefer-ences, when measured as concrete attitudesregarding specific issues, also can be goodpredictors of grades, above and beyond theeffects of skills, habits, and styles. That ourindicators of concrete attitudes successfullypredict educational outcomes disputesOgbu's (1991b) claim that African Ameri-cans' responsesto survey questions representmere wishful thinking and Mickelson's(1990) argument that African Americans'

positive attitudes toward school have littleimpact on performance. The message forscholars is clear: Indicators of specific atti-tudes regardingdaily school experiences canbe meaningful predictors of school perfor-mance. Indeed, models lacking these indica-tors areprobablymisspecified.

There is also a lesson here for the debateover policy and the controversy in the gen-eral public regardingthe extent to which Af-rican Americans are

responsiblefor the ra-

cial gap in school performance.The opposi-tional culture argumenthas led some schol-ars to argue that the greatesthope for reduc-ing the gap in educational performanceliesin African Americans changing their valuesystem. Forexample, D'Souza (1995) writes,

[W]henhundredsof thousandsof blackmengatheron theMall nWashingtonoemphasize




atonement, elf-help, entrepreneurship,trongfamilies,andtakingresponsibilityor theirac-tions, the world is changing and there aregrounds or hope. (P. xxxi)

This argument s misplaced. If anything,Af-rican Americans maintain more pro-schoolvalues and are more likely to esteem theirhigh-achieving peers than are whites. WhatAfrican Americans lack, however, are thematerial conditions that foster the develop-mentof skills, habits, and styles rewardedbyteachers.Furthermore, t is unlikely thatAf-ricanAmericans will demonstrate these cul-tural skills at the same level as whites until

they enjoy comparable material conditions.Rather than admonishing AfricanAmericansto take responsibility for improving theirown conditions, therefore, the best way toreduce the gap in educational performance sto implement policies that reduce economicinequality and residential segregation.

Our conclusions are tempered somewhatby limitations in our data. We used the 1990wave of the NELS primarily because it hasmany good indicators of students' resistancetoward school, but this decision restrictedour analyses to high school sophomores andtherefore limits the generalizability of ourstudy. We are guardedly optimistic that thepatterns presented here, however, extend toolder and younger high school students. Weexplored several indicators of school resis-tance in the third wave of the NELS whenthe students were seniors in high school, andthe results were highly consistent with the

patternswe reporthere. And we went back-ward to the 1988 base-year data when thestudents were eighth graders. Again, the fewavailable indicators of students' resistancetoward school produced results parallelingthe ones presented here, suggesting that ourfindings are not simply a function of the se-lectivity of a sample of studentspersisting totheir sophomore year in high school.

Finally, this is an example of how an at-tractive

theorycan become so

captivatingthat scholars no longer demandthe usual de-gree of methodological rigor. Although em-pirical corroborationshould be expected ateach stage of theory development, it is espe-cially importantwhen a theory's major as-sumptions are no longer seriously ques-tioned. Our study highlights the advantagesof testing a theory generated and supported

mainly by studies using one data collectionmethod (in this case ethnographies) with datafrom an alternativedata collection method(inthis case survey data). Oppositional culture

proponents have been too quick to disavowsurvey research, seeing it as inherently inca-pable of measuring resistance to school-justas some survey researchers reject ethno-graphic work because of its limitedgeneralizability. Ratherthan arguing for theinherent superiority of any one data collec-tion method, scholarswill benefit from strug-gling to reconcile apparentlydiscrepantfind-ings.

James W. Ainsworth-Darnells a Ph.D. candi-date at TheOhio State University.His researchinterests nclude tratification,acial/ethnic ela-tions,and sociology of education.His disserta-tion examineshowneighborhoodontextaffectsschoolperformance cross race and class. WithVincentRoscignohe is studyingracialvariationin educational eturnsor culturalcapital; abormarkets nd theeffectof vocational ducation noccupational trajectories; and (with VincentRoscignoand Maria VWlez)he impactof bilin-

gual educationon minority chievement.He hasan articleforthcoming in The Journalof Marriageand the Family (with Douglas B. DowneyandMikaela Dufur) on children's well-being insingle-parent ouseholds.

Douglas B. Downey s AssistantProfessorof So-ciology at The Ohio State University.His re-searchfocuses on stratificationprocesses withspecial attentionto race and ethnicityand thefamily. Currentlyhe is examining the conse-quencesof student-teacheracialcongruence n

teachers' ubjective valuations fstudentbehav-ior. Also, with Stefanie Neubauer and PhilipHostetler he is testing implications of the re-source dilutionexplanationor the inverserela-tionshipbetween he number f siblingsand edu-cationaloutcomes.

REFERENCES

Alexander, Karl L, Doris R. Entwisle, andSamuel D. Bedinger. 1994. When Expecta-tions Work: Race and Socioeconomic Differ-

ences in School Performance. Social Psychol-ogy Quarterly 57:283-99.

Bourdieu, Pierre. 1977. Reproduction in Educa-tion, Society, and Culture. Beverly Hills, CA:Sage.

Bowles, Samuel and Herbert Gintis. 1976.Schooling in Capitalist America. New York:Basic Books.

Cohen, Jacob and Patricia Cohen. 1983. Applied




Multiple Regression/Correlation Analysis forthe Behavioral Sciences. 2d ed. Hillsdale, NJ:Erlbaum.

Coleman, James S. 1961. The Adolescent Society.

New York: Free Press.Cose, Ellis. 1993. The Rage of a Privileged Class.New York: Harper Collins.

D'Souza, Dinesh. 1995. The End of Racism. NewYork: Free Press.

Dollard, John. 1957. Caste and Class in a South-ernTown. 3d ed. Garden City, NY: Doubleday.

Erickson, Frederick. 1987. TransformationandSchool Success: The Politics and Culture ofEducational Achievement. Anthropology andEducation Quarterly 18:335-57.

Farkas, George. 1996. Human Capital or Cultural

Capital? Ethnicity and Poverty Groups in anUrban School District. New York: Walter deGruyter.

Farley, Reynolds. 1984. Blacks and Whites:Nar-rowing the Gap. Cambridge, MA: HarvardUniversity Press.

Feagin, Joe R. 1991. The Continuing Signifi-cance of Race-Antiblack Discrimination inPublic Places. American Sociological Review56:101-16.

Fendrich,Michael and Connie M. Vaughn. 1994.

Diminished Lifetime Substance Use OverTime: An Inquiry Into Differential Under-reporting. Public Opinion Quarterly 58:96-123.

Fischer, Claude S. Michael Hout, Martin SanchezJankowski, Samuel R. Lucas, Ann Swidler, andKim Voss. 1996. Inequality by Design: Crack-ing the Bell Curve Myth. Princeton, NJ:Princeton University Press.

Foley, Douglas E. 1991. ReconsideringAnthro-pological Explanations of Ethnic School Fail-ure. Anthropology and Education Quarterly

22:60-81.Fordham, Signithia and John U. Ogbu. 1986.

Black Students' School Success: Coping withthe Burden of 'Acting White. ' The UrbanRe-view 18:176-206.

Gilmore, Perry. 1985. 'Gimme Room': SchoolResistance, Attitude, and Access to Literacy.Journal of Education 167:111-28.

Hochschild, Jennifer L. 1995. Facing Up to theAmerican Dream: Race, Class, and the Soul ofthe Nation. Princeton,NJ: PrincetonUniversityPress.

Jaynes, Gerald David and Robin M. Williams, Jr.1989. A Common Destiny: Blacks and Ameri-can Society.Washington, DC: National Acad-emy Press.

Jones, Elise F. and Jacqueline Darroch Forrest.1992. Underreportingof Abortion in Surveysof U.S. Women: 1976 to 1988. Demography29:113-26.

Kerckhoff, Alan C. and Richard T. Campbell.

1977. Black-White Differences in the Educa-tional Attainment Process. Sociology of Edu-cation 50:15-27.

Kohn, Melvin L. 1977. Class and Conformity:A

Study in Values, 2d ed. Chicago, IL: Univer-sity of Chicago Press.Lamont, Michelle and Annette Lareau. 1988.

Cultural Capital: Allusions, Gaps andGlissandos in Recent Theoretical Develop-ments. Sociological Theory 6:153-68.

MacLeod, Jay. 1995. Ain't No Makin' It: Aspira-tions and Attainment in a Low-Income Neigh-borhood. Boulder, CO: Westview Press.

Massey, Douglas S. and Nancy A. Denton. 1993.American Apartheid: Segregation and theMaking of the Underclass. Cambridge, MA:

HarvardUniversity Press.Mickelson, Roslyn Arlin. 1990. The Attitude-

Achievement Paradox among Black Adoles-cents. Sociology of Education 63:44-61.

Moynihan, Daniel P. 1965. TheNegro Family: ACase For National Action. Washington, DC:U.S. Government PrintingOffice.

National Center for Education Statistics. 1995.User's Manual: National Education Longitudi-nal Study of 1988. U.S. Departmentof Educa-tion, Office of Educational Research and Im-

provement.Ogbu, John U. 1974. The Next Generation: AnEthnography of Education in an Urban Neigh-borhood. New York: Academic Press.

. 1978. Minority Education and Caste.New York: Academic Press.

. 1991a. Low Performanceas an Adapta-tion: The Case of Blacks in Stockton, Califor-nia. Pp. 249-85 in Minority Status andSchooling, edited by M. A. Gibson and J. U.Ogbu. New York: GrandPublishing.

. 1991b. Minority Responses and School

Experiences. The Journal of Psychohistory18:433-56.

Porter, James R. 1974. Race, Socialization, andMobility in Early Occupational Attainment.American Sociological Review 39:306-16.

Portes, Alejandro and Kenneth L. Wilson. 1976.Black-White Differences in Educational At-

tainment. American Sociological Review 41:414-31.

Portes, AlejandroandMin Zhou. 1993. TheNewSecond Generation: Segmented Assimilationand Its Variants. The Annals

ofthe American

Academy PSS 530:74-96.Powdermaker, Hortense. 1968. After Freedom: A

Cultural Study in the Deep South. New York:Atheneum.

Solomon, R. Patrick. 1992. Black Resistance inHigh School. Albany, NY: State University ofNew York Press.

Solorzano, Daniel. 1991. Mobility Aspirationsamong Racial Minorities, Controlling for




SES. Sociology and Social Research 75:182-88.

Swidler, Ann. 1986. Culture n Action: Symbolsand Strategies. American Sociological Review

51:273-86.Tuan, Mia. 1995. Korean and Russian Studentsin a Los Angeles High School. Pp. 107-30 inCalifornia's Immigrant Children: Theory, Re-search, and Implications for EducationalPolicy, edited by R. G. Rumbaut and W. A.

Cornelius. San Diego, CA: University of Cali-fornia.

Weis, Louis. 1985. Between Two Worlds: BlackStudents in an Urban CommunityCollege. Bos-

ton, MA: Routledge and Kegan Paul.Willis, Paul. 1976. Learning to Labor. Lexington,MA: D.C. Heath.

Wilson, William J. 1996. When Work Disap-pears: The Worldof the New UrbanPoor. NewYork: Knopf.

assessing the opositional culture explanation for racial differences in school

Documents