Discrimination and Adjustment forMexican American Adolescents:

A Prospective Examination of the Benefitsof Culturally Related Values

Cady Berkel, George P. Knight, Katharine H. Zeiders,Jenn-Yun Tein, Mark W. Roosa, Nancy A. Gonzales, and Delia Saenz

Arizona State University

Mexican American adolescents face disparities in mental health and academic achievement,perhaps in part because of discrimination experiences. However, culturally related values, fos-tered by ethnic pride and socialization, may serve to mitigate the negative impact of discrim-ination. Guided by the Stress Process Model, the current study examined risk and protectiveprocesses using a 2-wave multi-informant study with 750 Mexican American families. Specifi-cally, we examined 2 possible mechanisms by which Mexican American values may supportpositive outcomes in the context of discrimination: as a protective factor (moderator) or riskreducer (mediator). Analyses supported the role of Mexican American values as a risk reducer.This study underscores the importance of examining multiple mechanisms of protective pro-cesses in understanding Mexican American adolescent resilience.

Latino adolescents in the United States are at elevated risk for a variety ofmental health and academic problems (Rumbaut, 1994). Mexican Americanadolescents, in particular, report higher levels of depressive symptoms (Hill,Bush, & Roosa, 2003) as well as externalizing behaviors (Gorman-Smith,Tolan, Henry, & Florsheim, 2000). Moreover, despite Mexican Americanfamilies’ emphasis on the value of school as an important pathway to suc-cess, Mexican American adolescents report lower academic self-efficacy(Fuligni, Witkow, & Garcia, 2005) and are more likely to drop out of school(Chavez, Oetting, & Swaim, 1994) than European American peers. Somehave suggested that these elevated mental health and academic risks are duein part to experiences of discrimination commonly reported by MexicanAmerican adolescents, which have been linked to poor mental health and

JOURNAL OF RESEARCH ON ADOLESCENCE, 20(4), 893–915r 2010 The AuthorsJournal of Research on Adolescence r 2010 Society for Research on AdolescenceDOI: 10.1111/j.1532-7795.2010.00668.x

Requests for reprints should be sent to Cady Berkel, Prevention Research Center, ArizonaState University, 900 S. McAllister Ave., Tempe, AZ 85287-6005. E-mail: cady.berkel@asu.edu

academic achievement (Coker et al., 2009; Stone & Han, 2005). Nevertheless,many Mexican American adolescents have good mental health outcomes, aresuccessful in school (Gonzales, Knight, Birman, & Sirolli, 2004), and oftenappear to be more resilient than one might expect given the stressors towhich they are exposed (Escobar, Nervi, & Gara, 2000). One potential mech-anism operating among Mexican American adolescents that may supportthis resilience is the socialization processes that lead to the internalization ofculturally related values associated with the ethnic group. Findings indicatethat Mexican American values directly relate to positive outcomes (Gonzaleset al., 2008) and can buffer or attenuate the negative effects of additionalstressors these youth experience (Umana-Taylor & Updegraff, 2007). Giventhis, the present study examined the potential for the socialization of cul-turally related values to mitigate the negative impact of discrimination on themental health and academic outcomes of Mexican American adolescents.

THE IMPACT OF DISCRIMINATION

Discrimination is a key stressor for members of minority groups, which maycontribute to persistent disparities across many important domains of ad-justment (e.g., Berkel et al., 2009; Coker et al., 2009; Greene, Way, & Pahl, 2006;Walls, Chapple, & Johnson, 2007; Wong, Eccles, & Sameroff, 2003). Nationalestimates indicate that 30% of Mexican American adults experience unfairtreatment daily due to their ethnicity; these estimates reached 50% for theyoungest cohort (18–24-year-olds) studied (Perez, Fortuna, & Alegrıa, 2008).Mexican American adolescents’ experiences of discrimination have beennegatively associated with both internalizing and externalizing mentalhealth problems (Greene et al., 2006; Romero, Martinez, & Carvajal, 2007;Romero & Roberts, 2003b; Smokowski & Bacallao, 2007) and academic out-comes (DeGarmo & Martinez, 2006; Stone & Han, 2005; Wayman, 2002).Because evidence of the impact of discrimination spans across many sig-nificant domains for adolescent adjustment, in both Mexican Americans andother minority groups, it could be a primary underlying cause of health andsocial disparities in this society. On the other hand, the expansiveness ofthese effects may be artificial, as previous research has generally limited itsfocus to a single outcome, and thus has not taken the possible biasing in-fluence of comorbidity into account, even though many of the outcomes areplainly correlated (e.g., Gorman-Smith et al., 2000; Manongdo & RamırezGarcıa, 2007). To reduce the possibility of spurious associations, a call hasbeen made to examine the general versus specific outcomes of psychosocialrisk factors like discrimination (McMahon, Grant, Compas, Thurm, & Ey,2003; Shanahan, Copeland, Costello, & Angold, 2008). Thus, in the currentstudy, we simultaneously examined the impact of discrimination onthe multiple domains of adjustment in Mexican American adolescents.Moreover, in a secondary analysis, we took an underutilized approach in

894 BERKEL ET AL.

modeling study outcomes as a latent variable to understand the effects ofdiscrimination on comorbidities and on specific outcomes.

BENEFITS OF MEXICAN AMERICAN VALUES

Although evidence for the impact of discrimination on Mexican Americanadolescent mental health and academic achievement is accumulating, westill know little about processes that support positive adjustment (Romero &Roberts, 2003b), despite the many Mexican American adolescents who aresuccessfully coping with discrimination (Edwards & Romero, 2008). Fre-quently noted is the protective nature of Mexican nativity for MexicanAmerican adolescents. Contrary to assumptions, adolescents born in Mexicohave lower rates of mental health problems than those born in the UnitedStates, a phenomenon that has become known as the ‘‘Mexican Americanparadox’’ (Escobar et al., 2000). Rather than nativity per se, the underlyingmechanism associated with the Mexican American paradox may actually bethe elements of culture that are attached to nativity; that is, those adolescentsoriginating in Mexico retain their ties to the protective elements of theirnative culture, while those born in the United States are less connected tosuch values (Gonzales et al., 2008).

The internalization of values rooted in Mexican culture, such as familism,respect for elders, and religiosity, are thought to enhance mental health andacademic achievement by nurturing strong ties, deference, and a sense ofobligation to family members and other prosocial adults, thus enabling theshaping of youth behavior in accordance with socially accepted conduct(Coohey, 2001; Gonzales et al., 2008). Such values have been associated withbetter academic grades and lower rates of internalizing and externalizingproblems among Latinos (Benjet et al., 2007; Coatsworth et al., 2002; Coon-rod, Balcazar, Brady, Garcia, & Van Tine, 1999; Esparza & Sanchez, 2008;Fuligni, Tseng, & Lam, 1999; Romero & Ruiz, 2007; Smokowski & Bacallao,2007; Unger et al., 2002). Gonzales et al. (2008) investigated whether theseculturally based values mediated the established relation between nativityand mental health and found that relations between nativity and both ex-ternalizing and academic engagement were indeed explained by decreasesin Mexican American values for U.S.-born adolescents. Thus, we suggest thatMexican American values are a source of resilience supporting academicachievement and mental health outcomes in contexts of discrimination.

Given the positive effects of Mexican American values, it is important toconsider how they may be maintained in a context marked by discriminationand acculturative pressure; the answer may lie in aspects of ethnic identityand familial ethnic socialization practices. High levels of ethnic identityamong Mexican American adolescents have been shown to be associatedwith increases in familism values over time (Baer & Schmitz, 2007). Impor-tantly, attachment to the ethnic group, including being proud of ethnic group

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 895

membership, may be a critical feature of ethnic identity for the endorsementof culturally related values among Mexican American adolescents becausethis sense of self as a member of the ethnic group makes the adolescentreceptive to the enculturative socialization efforts of the family (Armenta,Knight, Carlo, & Jacobson, in press). The contribution of socialization aboutcultural heritage (i.e., enculturation) to the development of ethnic identityhas been well established for several ethnic minority groups (Cheshire, 2001;Murry, Berkel, Brody, Miller, & Chen, 2009; Umana-Taylor, Bhanot, & Shin,2006). Parents’ socialization practices are thought to be critically importantfor fostering a sense of ethnic belonging, ethnic pride, and values in MexicanAmerican children and to protect them from potentially marginalizingexperiences such as discrimination (Umana-Taylor, Alfaro, Bamaca, &Guimond, 2009). Socialization messages that transmit culture, increasepride, and teach about equality are associated with positive outcomes forMexican American adolescents (Huynh & Fuligni, 2008), perhaps throughthe development of a sense of positive regard, affiliation, and pride for theircultural group (Kiang, Yip, Gonzales-Backen, Witkow, & Fuligni, 2006). Wesubmit that ethnic socialization messages will lay the foundation for thedevelopment of a strong sense of ethnic pride among adolescents, which willin turn enable them to retain their cultural values over time.

APPLICATION OF THE STRESS PROCESS MODEL

Because the study of the utility and functional nature of culturally relatedvalues is in its infancy, it remains unclear how Mexican cultural values mightreduce the likelihood of developing mental health and academic problems inthe face of repeated discrimination experiences. The Stress Process Model(Roosa, Wolchik, & Sandler, 1997; see Figure 1) emphasizes the importanceof understanding the sequence of events from stressor to outcome to thedevelopment of impactful preventive interventions. The model classifies

Discrimination(Stressor)

Externalizing, Internalizing, &Academic Achievement

(Mental Health Outcomes)

Mexican Cultural Values(Protective Factor)

(a)

Discrimination(Stressor)

Externalizing, Internalizing, &Academic Achievement

(Mental Health Outcomes)

Mexican Cultural Values(Risk Reducer)

(b)

FIGURE 1 Cultural Values as a Protective Factor (a) and as a Risk Reducer (b).

896 BERKEL ET AL.

processes as protective factors (moderators) or risk reducers (mediators). As aprotective factor, Mexican American values would attenuate the impact ofdiscrimination on mental health and academic outcomes. Umana-Taylor andUpdegraff (2007) found that Mexican American cultural orientations moder-ated the relations between discrimination and both depression and self-esteemfor boys, such that when traditional Mexican values were low, discriminationwas associated with higher depression and lower self-esteem. In contrast, whenvalues were high, discrimination was not significantly associated with depres-sion or self-esteem. These results suggest that cultural values of MexicanAmerican adolescents may be operating as protective factors by attenuating theeffects of stressors on negative outcomes. While this is the only study exam-ining the moderating role of Mexican American values on the influences ofdiscrimination, others (Finch, Kolody, & Vega, 2000) found that related culturalvariables, such as language and nativity, moderated the relations betweendiscrimination and depression among Mexican Americans. Furthermore,research with African Americans has demonstrated the moderating roles ofracial identity and racial socialization, in addition to cultural values, includingcommunity orientation and spirituality, in the association between discrimi-nation and mental health (Caldwell, Kohn-Wood, Schmeelk-Cone, Chavous, &Zimmerman, 2004; Fischer & Shaw, 1999; Scott, 2003; Wong et al., 2003).

Alternatively, the Stress Process Model suggests that protective processesmay function as risk reducers, which mediate the relation between a stressorand outcomes (Roosa et al., 1997). As a risk reducer, Mexican Americanvalues would have a positive effect on mental health and academic outcomes(Esparza & Sanchez, 2008; Smokowski & Bacallao, 2007). Further, if discrim-ination experiences promote racial/ethnic identity development as someauthors have suggested (Cross, Parham, & Helms, 1991; Pahl & Way, 2006),and if this greater endorsement of racial/ethnic identity fosters the internal-ization of culturally related values (Armenta et al., in press), discriminationmay encourage the internalization of Mexican American values that in turnpromote positive mental health and academic outcomes. Although limitedsupport for this perspective exists because possible associations betweendiscrimination and Mexican American values have been given scant atten-tion in the literature, Romero and Roberts (1998) found a positive relationbetween perceptions of discrimination and ethnic identity achievement(exploration). Another study found that discrimination was negatively cor-related with familism, although this path was not significant in the overallmodel testing multiple cultural influences on familism and externalizing(Smokowski & Bacallao, 2006). The current study is the first examination ofwhether cultural values act as a risk reducer, mediating the relations betweendiscrimination and mental health and academic outcomes for MexicanAmerican adolescents. We hypothesized that discrimination will result in anincrease in Mexican American values, which in turn will be associated withincreases in mental health and academic outcomes.

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 897

THE CURRENT STUDY

In the current prospective study, we sought to determine how discriminationand Mexican American values influence adolescent mental health and ac-ademic outcomes over time. We predicted that discrimination would be as-sociated with an increase in mental health and academic problems. However,the means by which positive cultural values might be beneficial for adoles-cent adjustment have not been established. Thus, we tested two possiblemechanisms, as suggested by the Stress Process Model: cultural values as amoderator and as a mediator. Moreover, previous research suggests thatethnic socialization lays the foundation for the development of ethnic prideand, in turn, may foster higher levels of protective Mexican American values(Umana-Taylor et al., 2009). Hence, we also considered how mothers’ ethnicsocialization and adolescents’ ethnic pride might serve to maintain positivecultural values despite the negative impact of acculturative pressures asso-ciated with discrimination experiences. Finally, we examined the specificand general effects of discrimination on mental health and academic out-comes by testing a model that controls for comorbidities.

METHOD

Participants

Data for this study come from the first and second waves of an ongoinglongitudinal studying investigating the role of culture and context in the livesof Mexican American families (Roosa et al., 2008). Participants were 750Mexican American students and their families who were selected from ros-ters of schools that served ethnically and linguistically diverse communitiesin a large southwestern metropolitan area. Eligible families met the followingcriteria: (a) they had a fifth grader attending a sampled school; (b) bothmother and child agreed to participate; (c) the mother was the child’s bi-ological mother, lived with the child, and self-identified as Mexican or Mex-ican American; (d) the child’s biological father was also reported to be ofMexican origin; (e) the child was not severely learning disabled; and (e) nostep-father or mother’s boyfriend was living with the child (unless the boy-friend was the biological father of the target child). Although participationwas optional for fathers, 460 (81%) fathers from the 570 two-parent familiesin the study also participated.

At Time 1, 22.9% were single-parent families and 77.1% were two-parentfamilies. In contrast to the majority of previous studies of Mexican Americanfamilies, this sample was diverse on both SES indicators and language(Roosa et al., 2008). Family incomes ranged from oUS$5,000 to 4US$95,000,with the average family reporting an income of US$30,000–US$35,000. Interms of language, 30.2% of mothers, 23.2% of fathers, and 82.5% of ado-lescents were interviewed in English. The mean age of mothers in our study

898 BERKEL ET AL.

was 35.9 (SD 5 5.81) and mothers reported an average of 10.3 (SD 5 3.67)years of education. The mean age of fathers was 38.1 (SD 5 6.26) and fathersreported an average of 10.1 (SD 5 3.94) years of education. The mean age ofadolescents (48.7% female) was 10.42 (SD 5 0.55). A majority of mothers andfathers were born in Mexico (74.3%, 79.9%, respectively), while a majority ofadolescents were born in the United States (70.3%).

At Time 2, approximately 2 years after Time 1 data collection, 711 familieswere re-interviewed, when most students were in the seventh grade. Of the39 families missing at Time 2, 16 refused to participate. Families who par-ticipated in Time 2 interviews were compared with families who did notparticipate in Time 2 interviews on several Time 1 demographic variablesand no differences emerged on child (i.e., gender, age, generational status,language of interview), mother (i.e., marital status, age, generational status),or father characteristics (i.e., age, generational status).

Procedure

Mothers (required), fathers (optional), and children (required) participatedin in-home Computer-Assisted Personal Interviews, scheduled at the fam-ily’s convenience. Interviews were about 2.5 hours long. Cohabitating familymembers’ interviews were conducted concurrently by trained interviewersin different locations at the participants’ home. Each interviewer received atleast 40 hours of training, which included information on project’s goals,characteristics of the target population, the importance of professional con-duct when visiting participants’ homes as well as throughout the process,and the critical role they would play in collecting the data. Interviewers readeach survey question and possible response aloud in participants’ preferredlanguage to reduce problems related to variations in literacy levels. Familieswere compensated US$45 and US$50 per participating family member atTime 1 and Time 2, respectively.

Measures

Discrimination. Adolescents responded to nine items designed to assessdiscrimination experiences from peers and teachers. Because no measure ofdiscrimination for Mexican Americans was available at the time of thisstudy’s development, three measures that had been validated for othergroups (Hughes and Dodge, 1997, Racism in the Workplace Scale; Landrineand Klonoff, 1996, Schedule of Racist Events; and Klonoff and Landrine,1995, Schedule of Sexist Events) were adapted to our population. The fourpeer items (e.g., ‘‘How often have kids at school called you names becauseyou are Mexican American?’’) and five teacher items (e.g., ‘‘How often haveyou had to work harder in school than White kids to get the same praise or

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 899

the same grades from your teachers because you are Mexican American?’’)relied upon a Likert-type response scale ranging from 1 5 not at all true to5 5 very true. Items reflected both personal experiences of discrimination andpublic regard. Cronbach’s a was .78 at both Time 1 and Time 2.

Mexican American values. The Mexican American Cultural Values Scale(Knight et al., 2010) was used to assess Mexican American values. The scalewas developed based upon focus groups conducted with Mexican Americanmothers, fathers, and adolescents about Mexican American and AngloAmerican cultures. The current study used five subscales from this measureto assess adolescents’ Mexican American values: supportive and emotionalfamilism (six items, e.g., ‘‘parents should teach their children that the familyalways comes first’’); obligation familism (five items, e.g., ‘‘if a relative ishaving a hard time financially, one should help them out if possible’’); referentfamilism (five items, e.g., ‘‘a person should always think about their familywhen making important decisions’’); respect (eight items, e.g., ‘‘childrenshould always be polite when speaking to any adult’’); and religiosity (sevenitems, e.g., ‘‘parents should teach their children to pray’’). Adolescentsresponded to items using a Likert-type scale ranging from 1 5 not at all to4 5 very much. Cronbach’s a was .87 at Time 1 and .90 at Time 2 (for furtherdetails about the results of CFA and the Cronbach’s as by broken down bysubscale and total familism, see Knight et al., 2010).

Ethnic socialization. Ethnic socialization was assessed with anadaptation of the 10-item Ethnic Socialization Scale from the EthnicIdentity Questionnaire developed by Bernal and Knight (1993). Usingmothers’ report, the scale assesses the extent to which mothers socializedchildren into Mexican culture. A sample item asked how often mothers ‘‘tell[target child] that the color of a person’s skin does not mean that person isbetter or worse than anyone else’’ and ‘‘take [target child] to Mexicancelebrations like Quincieneras, Mexican weddings, or baptisms.’’ Responsesranged from 1 5 almost never or never to 4 5 a lot of the time/frequently.Cronbach’s a was .76.

Ethnic pride. A four-item scale, developed for the current study andpilot tested with a sample of 162 Mexican American adolescents (Thayer,Valiente, Hageman, Delgado, & Updegraff, 2002), assessed ethnic pride forMexican Americans. Adolescents were asked to rate how much they agreedor disagreed with each item (e.g., ‘‘You feel proud to see Latino actors,musicians and artists being successful’’) with responses ranging from 1 5 notat all true to 5 5 very true. Cronbach’s a was .71.

Adolescent symptoms of conduct disorder (CD), oppositional defiantdisorder (ODD), anxiety, and depression. We used the computerized

900 BERKEL ET AL.

version of the Diagnostic Interview Schedule for Children (DISC-IV), astructured diagnostic instrument designed for use by nonclinicians, to assesschild mental health symptomatology (Shaffer, Fisher, Lucas, Dulcan, &Schwab-Stone, 2000). The DISC-IV can be used to assess psychiatricsymptoms/diagnoses that occur in children and adolescents using DSM-IV (American Psychiatric Association, 1994) and ICD-10 (World HealthOrganization, 1993) criteria. The DISC was originally developed in 1979 andhas been successfully translated into Spanish according to psychometricwork done in Puerto Rico (Bravo et al., 2001). Adolescents respond to a seriesof modules that assess symptom counts. For the current study we relied oncounts of the following symptoms: CD, ODD, anxiety, and depression.

Academic self-efficacy. Adolescents’ reports on academic self-efficacywere assessed using the Patterns of Adaptive Learning Survey (Midgley,Maehr, & Urdan, 1996). Items are not specific to subject matter or tasks, butthey are specific to students’ classroom experiences (e.g., ‘‘I am certain I canmaster the skills taught in school this year’’). Students respond to the sixitems with options ranging from 1 5 not at all true to 5 5 very true. Cronbach’sas were .75 and .79 at Time 1 and Time 2, respectively.

Academic achievement. To assess academic performance, we relied onteacher reports. Specifically, math and English teachers were asked, ‘‘If youwere giving final grades today, what grade would this student receive inyour course?’’ Teachers responses (1) A to (5) E/F were averaged to computeone score. This academic achievement outcome score was only available inthe Time 2 assessments.

Analytic plan. Analyses were conducted using SEM with the fullinformation maximum likelihood estimation procedures available in MplusVersion 5.1 (Muthen & Muthen, 2008), unless otherwise noted. In all models,discrimination (peer and teacher subscales), Mexican American values(familism-support, familism-obligation, familism-referent, respect, andreligiosity subscales), internalizing (anxiety and depression), and external-izing (CD and ODD) were treated as latent variables. The remainingvariables (ethnic socialization, ethnic pride, academic self-efficacy, andgrades) were treated as observed variables.

Because cross-sectional data have led to controversies in the literature ondirection of effects for discrimination and outcomes such as mental healthand cultural affiliation (e.g., Phinney, Madden, & Santos, 1998; Romero &Roberts, 2003a; Ruggiero, Taylor, & Lambert, 1996), it was necessary to con-duct preliminary analyses to evaluate the direction of these effects. We con-ducted a series of four cross-lag models to evaluate the direction of effectsbetween discrimination and (1) Mexican American values, (2) internalizing,(3) externalizing, and (4) academic self-efficacy. Because grades were not

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 901

collected at Time 1 we did not conduct a cross-lag model between discrim-ination and grades. However, because adolescents’ teachers were different inTime 1 and Time 2, we would not expect that grades at Time 1 would in-fluence discrimination at Time 2. For each of the other four variables relatedto discrimination, we compared a model in which the cross-lag paths wereconstrained to be equal to a model in which the cross-lag paths were free tovary. A significant difference in model fit would indicate that the strengths ofthe cross-lag relations were unequal and suggest the direction of causality(Brody et al., 2006). In addition, because nativity may be related to thesevariables, we examined all of these associations by child birthplace (UnitedStates or Mexico), using multiple group analyses in Mplus. If nativity wasfound to moderate any of the associations, we planned to include nativity inthe main analyses.

In the two main analysis models assessing the beneficial role of MexicanAmerican values, i.e., as a protective factor (i.e., moderation) or as a riskreducer (i.e., mediation), multiple fit indices (CFI, RMSEA, and SRMR) wereused to evaluate the extent to which the model fit the data, because no singleindicator is unbiased in all analytic conditions. We considered good (accept-able) model fit as indicated by a SRMR � .05 (.08) and either a RMSEA � .05(.08) or a CFI� .95 (.90) because simulation studies revealed that using thiscombination rule resulted in low Type I and Type II error rates (Hu & Bentler,1999). Time 1 levels of the outcome variables (except for teacher reports ofadolescents’ grades) were included in both models to enable the prediction ofchange over time. In addition, because academic self-efficacy is a knownpredictor of achievement (Buriel, Perez, De Ment, Chavez, & Moran, 1998),we also included a path from academic self-efficacy to grades for bothmodels.

In the ‘‘protective factor’’ model testing moderation effects, we created aninteraction with the discrimination and Mexican American values latentvariables at Time 1 (Klein & Moosbrugger, 2000). A significant direct pathfrom the interaction term to the outcomes would indicate the presence ofmoderation and would be probed by a multiple group analysis examiningthe relations between discrimination and outcomes at different levels ofMexican American values. In the ‘‘risk reducer’’ model testing mediation, weused bias-corrected bootstrap confidence intervals to assess the significanceof the standardized indirect effects (MacKinnon, Lockwood, Hoffman, West,& Sheets, 2002; Taylor, MacKinnon, & Tein, 2008). Mediation is consideredsignificant when the 95% CI does not encompass zero.

Finally, given the importance of unraveling the potential bias introducedby comorbidity effects and that similar studies show variables addressed inthis paper to be highly correlated (Lewinsohn, Zinbarg, Seeley, Lewinsohn, &Sack, 1997), the current study applied latent variable structural models toseparate specific from general effects as suggested by Bentler (1990). In thisapproach, comorbidity is treated as a single latent construct, with adolescent

902 BERKEL ET AL.

mental health outcomes (anxiety, depression, CD, and ODD) as indicators.The path from the predictors to the latent construct represents the effect onthe comorbidity of the disorders; additional paths from the predictors to eachof the indicators represent predictors’ effects on the specific disorders, afteraccounting for comorbidity.

RESULTS

Preliminary Analyses

Table 1 presents the correlations, means, and standard deviations for allstudy variables, calculated in SPSS 16.0. In general, experiences of discrim-ination were associated with poorer mental health and academic outcomes.Ethnic socialization was positively associated with ethnic pride. Adolescentswith high levels of ethnic pride also highly endorsed Mexican Americanvalues and reported more positive academic outcomes. Mexican Americanvalues at Time 2 were associated with lower externalizing symptoms. Finally,the mental health and academic outcomes were related, with externalizingand internalizing symptoms positively associated with one another andnegatively associated with academic outcomes.

In Figure 2, we present the results of the first of a series of cross-lag modelsin which we compared the cross-lag pathways between discrimination andMexican American values. All factor loadings were significant and above .50,indicating the adequacy of the measurements. The unconstrained model fitsignificantly better than the model in which the two paths were constrainedto be equal (Dw2 5 3.94, df 5 1, p � .05). This indicated that the standardized,unconstrained path from Time 1 discrimination to Time 2 Mexican Americanvalues (b5 .15, p � .001) was significantly stronger than the path from Time1 Mexican American values to Time 2 discrimination (b5 .00, ns), suggestingthat discrimination was likely producing changes in Mexican American val-ues, rather than the reverse. Similar preliminary cross-lag analyses wereconducted for internalizing, externalizing, and academic self-efficacy. Forinternalizing, the change in w2 (Dw2 5 9.25, df 5 1, p � .01) was again signifi-cant, indicating a worse fit of the constrained model, with the path fromdiscrimination to internalizing (b5 .19, p � .01) significantly stronger thanthe path from internalizing to discrimination (b5 � .12, ns), which was notsignificant. For externalizing, the change in w2 (Dw2 5 0.04, df 5 1, p4.10),indicating that the paths were equal in magnitude. Finally, for academic self-efficacy, the change in w2 (Dw2 5 9.66, df 5 1, p � .01) between the nestedmodel indicates worse fit, again with the path from discrimination to aca-demic self-efficacy (b5 � .13, p � .01) significantly stronger than the oppo-site path from self-efficacy to discrimination (b5 � .02, ns). Thus, in all cases,except externalizing, where the paths were equal, Time 1 discrimination wasthe driver of the relations with Time 2 outcomes. Next, we conducted a

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 903

TA

BL

E1

Co

rrel

atio

ns,

Mea

ns,

and

Sta

nd

ard

Dev

iati

on

sfo

rA

llS

tud

yV

aria

ble

s

12

34

56

78

910

1112

1314

1.D

iscr

imin

atio

nT

1—

2.D

iscr

imin

atio

nT

2.3

3nnn

—

3.E

thn

icso

cial

izat

ion

T1

.02

.071

—

4.E

thn

icp

rid

eT

1�

.04

.03

.13n

nn

—

5.E

thn

icp

rid

eT

2�

.02

.00

.19n

nn

.29n

nn

—

6.M

exic

anA

mer

ican

val

ues

T1

.01

.05

�.0

3.5

2nnn

.14n

nn

—

7.M

exic

anA

mer

ican

val

ues

T2

.08n

.04

.03

.18n

nn

.41n

nn

.32n

nn

—

8.E

xte

rnal

izin

gT

1.1

4nnn

.13n

nn�

.01

�.0

2�

.03

�.0

5�

.10n

nn

—

9.E

xte

rnal

izin

gT

2.0

61.2

8nnn

.00

.04

�.0

5.0

4�

.18n

nn

.39n

nn

—

10.

Inte

rnal

izin

gT

1.3

0nnn

.16n

nn

.01

�.0

8n�

.02

.01

�.0

1.3

0nnn

.15n

nn

—

11.

Inte

rnal

izin

gT

2.2

2nnn

.27n

nn�

.03

�.0

4�

.04

.01

�.0

2.1

8nnn

.41n

nn

.41n

nn

—

12.

Aca

dem

icse

lf-e

ffic

acy

T1

�.0

7n�

.04

.04

.38n

nn

.14n

nn

.40n

nn

.26n

nn�

.14n

nn�

.071

�.1

1nn�

.08n

—

13.

Aca

dem

icse

lf-e

ffic

acy

T2

�.1

2nn�

.10n

.04

.20n

nn

.27n

nn

.14n

nn

.41n

nn�

.11n

n�

.17n

nn�

.071

�.1

1nn

.40n

nn

—

14.

Gra

des

�.1

7nnn�

.17n

nn

.03

.13n

nn

.071

.05

�.0

1�

.14n

nn�

.13n

nn�

.071

�.1

2nn

.17n

nn

.26n

nn

—

Mea

ns

14.9

13.2

17.3

4.3

4.5

140.

013

5.0

1.7

2.7

17.7

14.4

26.2

25.3

2.6

Sta

nd

ard

dev

iati

on

s5.

84.

83.

90.

70.

610

.211

.92.

43.

212

.010

.13.

43.

81.

2

Not

e.E

xte

rnal

izin

g:

con

du

ctd

iso

rder

and

op

po

siti

on

ald

efia

nt

dis

ord

er;

inte

rnal

izin

g:

dep

ress

ion

and

anx

iety

.

T1,

Tim

e1;

T2,

Tim

e2.

1p�

.10;

np�

.05;

nnp�

.01;

nnnp�

.001

.

904 BERKEL ET AL.

multiple group analysis to examine child nativity as a possible moderator ofthe relations in each of these models. Given that we found no differences inassociations for those born in the United States and those born in Mexico, wedid not include nativity in either the protective factor or risk reducer models.

Test of Protective Factor Effects

To test whether Mexican American values served as a protective factor, weexamined a model in which these culturally based values moderated therelations between adolescent experiences of discrimination and each of thestudy outcomes: internalizing, externalizing, academic self-efficacy, andteacher-reported grades. The model included direct paths from the discrim-ination and Mexican American values latent constructs to the adolescentoutcomes, from ethnic socialization and ethnic pride paths to MexicanAmerican values, and from the discrimination �Mexican American valuesinteraction term to adolescent adjustment. In addition, to incorporate theformation of Mexican American values as part of the protective process, themodel included paths from ethnic socialization to ethnic pride to MexicanAmerican values. We also controlled for Time 1 levels of internalizing, ex-ternalizing, and academic self-efficacy (grades were not collected at Time 1).Fit indices did not support the role of Mexican American values as a pro-tective factor, w2(536) 5 2418.72, p � .01; CFI 5 .76; and RMSEA 5 .07. Fur-ther, none of the pathways from the interaction construct to outcomes weresignificant, indicating that Mexican American values did not moderate (i.e.,serve as a protective factor) the relation between discrimination and declinesin mental health or academic outcomes.

Test of Risk Reducer Effects

To test whether Mexican American values served as a risk reducer, we ex-amined a model in which these culturally based values mediated the

Discrimination

Mexican Cultural Values

Time 1

Discrimination

Mexican Cultural Values

Time 2

FIGURE 2 Discrimination and Mexican Cultural Values: Direction of Effects.

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 905

relations between discrimination and changes in adolescent outcomes(see Figure 3). The model included direct paths from Time 1 discriminationto the proposed study outcomes at Time 2 (internalizing, externalizing, ac-ademic self-efficacy, and teacher-reported grades), controlling for Time 1outcomes (except for teacher-reported grades), and mediated paths fromTime 1 discrimination to Mexican American values at Time 2 (controlling forTime 1 Mexican American values) to the outcomes at Time 2. As described inthe protective factor model, we incorporated paths from ethnic socializationto ethnic pride to Mexican American values. Fit indices demonstrated sup-port for the model, w2(238) 5 666.82, p � .001; RMSEA 5 .05 (.04–.05);CFI 5 .93; and SRMR 5 .05.

Standardized coefficients demonstrated that Mexican American valuesproduced a compensating effect for the negative impact of discriminationexperiences on mental health and academic outcomes. In addition, both thesocialization pathways and discrimination were independently associatedwith an increase in Mexican American values from Time 1 to Time 2. MexicanAmerican values, in turn, predicted improvements in internalizing, external-izing, and academic self-efficacy (controlling for respective Time 1 scores). Thetest of the mediated effects using bias-corrected bootstrap confidence intervals(MacKinnon et al., 2002) were significant for externalizing (95% CI: � .05,� .002), internalizing (95% CI: � .03, � .004), academic self-efficacy (95% CI:.01, .09), and grades (95% CI: .002, .03), supporting the role of Mexican culturalvalues as a risk reducer. As expected, ethnic socialization at Time 1 predictedethnic pride at Time 2, which was associated with a change in Mexican

FIGURE 3 A Test of Mexican American Values as a Risk Reducer.

906 BERKEL ET AL.

American values from Time 1 to Time 2. The mediated effect of ethnic pridealso was significant (95% CI: .01, .02). For example, to calculate the mediatedeffect from discrimination to academic self-efficacy, through Mexican Amer-ican values, one would multiply the coefficient of the path between discrim-ination and Mexican American values (i.e., .13) by the coefficient of the pathbetween Mexican American values and self-efficacy (i.e., .40), producing amediated effect of .05. The direct effect of discrimination on self-efficacy is� .19. Because of the inconsistent mediation (i.e., the combination of positiveand negative paths coefficients), combining the direct and mediated effectsresults in a total effect (� .14) that is smaller in magnitude than the direct effect.In this case, the mediated effect of Mexican American values reduces thenegative effects of discrimination by roughly one quarter of the direct effect.Also note that this mediation represents the enhancement of each relatedoutcome independent of the positive effects of the socialization effects.

Results of the comorbidity analyses showed that the model fit the dataadequately, w2(248) 5 673.76, p � .001; RMSEA 5 .05 (.04–.05); CFI 5 .92; andSRMR 5 .06. Time 2 Mexican American values had a significant negativeeffect on the comorbidity factor (b5 � .15, p � .005), but the effect of Time 1discrimination on the comorbidity factor was not significant (b5 .06, ns). Thepath coefficients between the other predictors, mediators, and outcomeswere consistent with those shown in Figure 3, with at most .01 to .02 differ-ences. After accounting for comorbidity, we examined the specific effectsfrom Time 1 discrimination and Time 2 Mexican American values to the twointernalizing indicators (anxiety and depression) and the specific effects tothe two externalizing indicators (CD and ODD symptoms). As Bentler (1990)suggested, to avoid the parameter identification issue, we examined thespecific effects to internalizing and to externalizing indicators separately. Theresults showed that the experience of discrimination, but not MexicanAmerican values, had specific effects on anxiety and depression (b5 .27,p � .01, b5 .11, p � .05, respectively). On the other hand, Mexican Americanvalues, but not the experience of discrimination, had specific effects on CDand ODD (b5 � .15, p � .001, b5 � .09, p � .05, respectively).

DISCUSSION

Despite a large and growing population of Mexican Americans in the UnitedStates in general and the southwestern states in particular (U.S. CensusBureau, 2008), discrimination endures, as is evident in reports of racial pro-filing by some police (Gonzales, 2008) and an active civilian border militiawith the motto ‘‘come get some’’ (Holthouse, 2005). Because such discrim-ination still exists, and may contribute to disparities across health and socialdomains for Mexican Americans, it is critical to understand the risk pro-cesses underlying the experience of discrimination. Furthermore, given thepotential impact of such disparities, a better understanding of the potential

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 907

protective processes that may alleviate these risks could provide valuableinformation for the development of preventive interventions to mitigate thenegative effects of stressors such as discrimination (Roosa et al., 1997).

Consistent with the developing literature (DeGarmo & Martinez, 2006;Gee, Ryan, Laflamme, & Holt, 2006), we found a strong link between expe-riences with discrimination and negative mental health and academic out-comes in a diverse and relatively representative sample of MexicanAmerican adolescents. However, the previous literature has provided mixedinterpretations of the direction of this relation, with some authors suggestingthat those with poor mental health are more likely to perceive ethnicity-related discrimination (e.g., Phinney et al., 1998) and others suggesting thatdiscrimination experiences lead to negative mental health and academicoutcomes (e.g., DeGarmo & Martinez, 2006; Umana-Taylor & Updegraff,2007). Unfortunately, this developing literature has been based almost ex-clusively upon the examination of the relations between discrimination andmental health and academic outcomes at a single point in time and, as such,has not permitted analyses that could provide evidence about the causaldirection of this relation. Results from our prospective, cross-lag analysesprovided some empirical support for the suggestion that discriminationleads to worsening mental health and academic achievement over time. Inaddition, results from our comorbidity analyses suggest that discriminationhas specific and unbiased effects on these outcomes.

Seeking to discover sources of resilience among Mexican American ad-olescents in the context of discrimination, we considered the protective in-fluence of the ethnic socialization and the internalization of MexicanAmerican values (Gonzales et al., 2004). Evidence is beginning to accumu-late that culturally related values are important for adolescent mental healthand academic outcomes (Gonzales et al., 2008; Smokowski & Bacallao, 2007).Indeed, we found that adolescents who experience more ethnic socializationfrom their parents report more pride in their ethnic group and greater en-dorsement of Mexican American values. In turn, these Mexican Americanvalues were associated with a reduced risk of negative mental health symp-toms over time and increases in academic self-efficacy. Further, this increasein academic self-efficacy was associated with higher teacher-reportedgrades. The familism, respect, and religiosity values assessed by our indexof Mexican American values may foster resilience by promoting strongbonds to the family and other community members (Gonzales et al., 2008).Indeed, Armenta et al. (in press) found that familism values explained thelink between ethnic identity and prosocial behaviors toward others. Thisinterconnectedness may serve to regulate adolescents’ acting out behaviorsin response to stressors. That these values are more associated with behav-ioral manifestations, rather than alleviating internal symptoms of depressionand anxiety, has important implications for the study of specificity of pre-dictors on outcomes and should be examined further in future studies.

908 BERKEL ET AL.

Our findings provide support for Mexican American values as a risk re-ducer, rather than a protective factor, to counteract the effects of discrim-ination on internalizing and academic outcomes. That is, in addition to thesocialization effects leading to the greater endorsement of Mexican Americanvalues, adolescents who experienced more discrimination also endorsedMexican American values to a greater extent, which in turn was associatedwith positive changes in mental health and academic self-efficacy. The lack ofsupport for moderation in the current study is in contrast to previous re-search with Mexican American and other ethnic minority groups (Caldwellet al., 2004; Fischer & Shaw, 1999; Scott, 2003; Umana-Taylor & Updegraff,2007; Wong et al., 2003). However, there are several features of our researchthat strongly support the validity and generalizability of our findings. First,our findings are based on a very diverse, representative, and stratified ran-dom sample of Mexican American families from a large metropolitan area.We conducted a stratified sampling of schools and randomly sampled stu-dents within those schools. In addition, the longitudinal, prospective designof our study allowed us to model changes in mental health and academicoutcomes over a substantial time period. Moreover, concurrent moderationis not necessarily inconsistent with the longitudinal changes in mental healthand academic outcomes being mediated by changes in Mexican Americanvalues.

Our findings are consistent with theories suggesting that cultural affili-ation is a coping mechanism prompted by experiences with discrimination.That such adaptation occurs in the context of a major stressor fits well withinclassic definitions of resilience (Luthar, Cicchetti, & Becker, 2000), such thatencountering the stressor leads to more positive outcomes than would beexpected. Moreover, the finding that adolescents can respond resilientlywhen confronted by discrimination has been echoed in research with AfricanAmerican adolescents (Berkel et al., 2009), where incidents that produceawareness about race and ethnicity are the primary means by which identitydevelopment is thought to occur (Cross et al., 1991). Discrimination eventsoften lead adolescents to explore their group membership (Pahl & Way,2006). As a result, they may become further embedded and take pride in theirethnic communities, which may foster the emic values that mitigate thenegative effects of discrimination. Although the current prospective designgoes beyond previous research in the ability to test directional hypotheses,future research should track trajectories of discrimination, Mexican Amer-ican values, and adolescent adjustment over a greater time period to enablemore definitive conclusions about casual directions.

This study emphasized the importance of Mexican American values inreducing the negative influence of discrimination for adolescents, which hasimportant implications for both prevention and policy. Related to preventionprogramming, the present findings suggest that programs designed to im-prove the health and academic outcomes of Mexican American adolescents

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 909

should include modules that fortify Mexican American values. In addition,because of the primacy of the family in the socialization of these values, itwould also follow that family-based prevention may be especially relevant inthis population (Murry, Berkel, Pantin, & Prado, 2010). Parenting programsseeking to encourage resilience among adolescents, moreover, should alsoinclude an emphasis on the importance of parents practicing ethnic social-ization with their children. Furthermore, prevention programs for margin-alized groups must invest high levels of time and resources into undoing thenegative effects of racism (e.g., Murry, Berkel, Brody, Gerrard, & Gibbons,2007). A more efficient and equitable approach might be large-scale pro-grams to reduce the prejudice and discrimination that maintain disparitiesamong children and families.

Relating to the policy domain, many regulations have been developedunder the assumption that socializing minority children into the majorityculture while discouraging the retention of traditional values and lifestyles isthe best way to support their well-being and the well-being of society. How-ever, efforts to encourage the use of English at the cost of Spanish (Barkeret al., 2001) and to adopt mainstream U.S. values, for example, may inad-vertently push children away from the cultural values that appear to beprotecting them (Gonzales et al., 2008). On the other hand, while we presentsolid evidence for the negative impact of teacher discrimination, there are nolarge-scale training efforts to enable teachers to appreciate diversity andincorporate cultural strengths into curriculums, thereby reducing this sourceof risk. In fact, policies that limit Spanish speaking in schools may actuallyincrease the discrimination to which adolescents and their parents are ex-posed. As a result, the current political climate associated with the immi-gration of persons from Mexico and other nearby countries appears to beincreasing the likelihood that Mexican Americans will experience discrim-ination, a risk factor, while simultaneously reducing the beneficial mecha-nisms such as traditional cultural values for adolescents. The continuation ofsuch polices likely would perpetuate the Mexican American paradox andshould not be supported in light of the demonstrated importance of cultur-ally based values in reducing risks to Mexican American youth as a result ofdiscrimination.

ACKNOWLEDGMENTS

Work on this paper was supported, in part, by grant MH68920 (Culture,context, and Mexican American mental health) and the Cowden Fellowshipprogram of the School of Social and Family Dynamics at Arizona State Uni-versity. The first author’s involvement in the preparation of this manuscriptwas supported by Training Grant T32MH18387. The authors are thankful forthe support of Marisela Torres, Jaimee Virgo, our Community AdvisoryBoard and interviewers, and the families who participated in the study.

910 BERKEL ET AL.

REFERENCES

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4thed.). Washington, DC: Author.

Armenta, B. E., Knight, G. P., Carlo, G., & Jacobson, R. P. (in press). The relation between ethnicgroup attachment and prosocial tendencies: The mediating role of cultural values. EuropeanJournal of Social Psychology.

Baer, J. C., & Schmitz, M. F. (2007). Ethnic differences in trajectories of family cohesion forMexican American and non-Hispanic White adolescents. Journal of Youth and Adolescence, 36,583–592.

Barker, V., Giles, H., Noels, K., Duck, J., Hecht, M., & Clement, R. (2001). The English-onlymovement: A communication analysis of changing perceptions of language vitality. Journal ofCommunication, 51, 3–37.

Benjet, C., Borges, G., Medina-Mora, M. E., Fleiz, C., Blanco, J., Zambrano, J., et al. (2007).Prevalence and socio-demographic correlates of drug use among adolescents: Results fromthe Mexican Adolescent Mental Health Survey. Addiction, 102, 1261–1268.

Bentler, P. M. (1990). Latent variable structural equation modeling in health services research. InL. Sechrest, E. Perrin, & J. Bunker (Eds.), Research methodology: Strengthening causal interpre-tations of nonexperimental data (pp. 61–83). Washington, DC: U.S. Department of Health andHuman Services.

Berkel, C., Murry, V. M., Hurt, T. R., Chen, Y.-F., Brody, G. H., Simons, R. L., et al. (2009). It takes avillage: Protecting rural African American youth in the context of racism. Journal of Youth andAdolescence, 38, 175–188.

Bernal, M. E., & Knight, G. P. (Eds.). (1993). Ethnic identity: Formation and transmission amongHispanics and other minorities. Albany, NY: State University of New York Press.

Bravo, M., Ribera, J., Rubio-Stipec, M., Canino, G., Shrout, P., & Ramırez, R., et al. (2001). Test-retest reliability of the Spanish version of the Diagnostic Interview Schedule for Children(DISC-IV). Journal of Abnormal Child Psychology, 29, 433–444.

Brody, G. H., Chen, Y.-F., Murry, V. M., Ge, X., Simons, R. L., Gibbons, F. X., et al. (2006). Perceiveddiscrimination and the adjustment of African American youths: A five-yearlongitudinal analysis with contextual moderation effects. Child Development, 77, 1170–1189.

Buriel, R., Perez, W., De Ment, T. L., Chavez, D. V., & Moran, V. R. (1998). The relationship oflanguage brokering to academic performance, biculturalism, and self-efficacy among Latinoadolescents. Hispanic Journal of Behavioral Sciences, 20, 283–297.

Caldwell, C. H., Kohn-Wood, L. P., Schmeelk-Cone, K. H., Chavous, T. M., & Zimmerman, M. A.(2004). Racial discrimination and racial identity as risk or protective factors for violent behaviorsin African American young adults. American Journal of Community Psychology, 33, 91–105.

Chavez, E. L., Oetting, E. R., & Swaim, R. C. (1994). Dropout and delinquency: Mexican-Amer-ican and Caucasian non-Hispanic youth. Journal of Clinical Child Psychology, 23, 47–55.

Cheshire, T. C. (2001). Cultural transmission in urban American Indian families. AmericanBehavioral Scientist, 44, 1528–1535.

Coatsworth, J. D., Pantin, H., McBride, C., Briones, E., Kurtines, W., & Szapocznik, J. (2002).Ecodevelopmental correlates of behavior problems in young Hispanic females. Applied De-velopmental Science, 6, 126–143.

Coker, T. R., Elliott, M. N., Kanouse, D. E., Grunbaum, J. A., Schwebel, D. C., Gilliland, M. J.,et al. (2009). Perceived racial/ethnic discrimination among fifth-grade students and its as-sociation with mental health. American Journal of Public Health, 99, 878–884.

Coohey, C. (2001). The relationship between familism and child maltreatment in Latino andAnglo families. Child Maltreatment, 6, 130–142.

Coonrod, D. V., Balcazar, H., Brady, M. J., Garcia, S., & Van Tine, M. (1999). Smoking,acculturation and family cohesion in Mexican-American women. Ethnicity and Disease, 9,434–440.

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 911

Cross, W. E., Parham, T. A., & Helms, J. E. (1991). The stages of Black identity development:Nigrescence models. In R. L. Jones (Ed.), Black psychology (3rd ed., pp. 319–338). Berkeley, CA:Cobb & Henry Publishers.

DeGarmo, D. S., & Martinez, C. R. (2006). A culturally informed model of academic well-beingfor Latino youth: The importance of discriminatory experiences and social support. FamilyRelations, 55, 267–278.

Edwards, L. M., & Romero, A. J. (2008). Coping with discrimination among Mexican descentadolescents. Hispanic Journal of Behavioral Sciences, 30, 24–39.

Escobar, J. I., Nervi, C. H., & Gara, M. A. (2000). Immigration and mental health: MexicanAmericans in the United States. Harvard Review of Psychiatry, 8, 64–72.

Esparza, P., & Sanchez, B. (2008). The role of attitudinal familism in academic outcomes: A studyof urban, Latino high school seniors. Cultural Diversity and Ethnic Minority Psychology, 14, 193–200.

Finch, B. K., Kolody, B., & Vega, W. A. (2000). Perceived discrimination and depression amongMexican-origin adults in California. Journal of Health and Social Behavior, 41, 295–313.

Fischer, A. R., & Shaw, C. M. (1999). African Americans’ mental health and perceptions of racistdiscrimination: The moderating effects of racial socialization experiences and self-esteem.Journal of Counseling Psychology, 46, 395–407.

Fuligni, A. J., Tseng, V., & Lam, M. (1999). Attitudes toward family obligations among Americanadolescents with Asian, Latin American, and European backgrounds. Child Development, 70,1030–1044.

Fuligni, A. J., Witkow, M., & Garcia, C. (2005). Ethnic identity and the academic adjustment ofadolescents from Mexican, Chinese, and European backgrounds. Developmental Psychology,41, 799–811.

Gee, G. C., Ryan, A., Laflamme, D. J., & Holt, J. (2006). Self-reported discrimination and mentalhealth status among African descendants, Mexican Americans, and other Latinos in the NewHampshire REACH 2010 Initiative: The added dimension of immigration. American Journal ofPublic Health, 96, 1821–1828.

Gonzales, D. (2008). Crime-sweep records raise suspicions of racial profiling. The Arizona Re-public, A1, A16–A17.

Gonzales, N. A., German, M., Kim, S. Y., George, P., Fabrett, F. C., Millsap, R., et al. (2008).Mexican American adolescents’ cultural orientation, externalizing behavior and academicengagement: The role of traditional cultural values. American Journal of Community Psychology,41, 151–164.

Gonzales, N. A., Knight, G. P., Birman, D., & Sirolli, A. A. (2004). Acculturation and enculturationamong Latino youth. In K. I. Maton, C. J. Schellenbach, B. J. Leadbeater, & A. L. Solarz (Eds.),Investing in children, youth, families, and communities: Strengths-based research and policy (pp.285–302). Washington, DC: American Psychological Association.

Gorman-Smith, D., Tolan, P. H., Henry, D. B., & Florsheim, P. (2000). Patterns of family func-tioning and adolescent outcomes among urban African American and Mexican Americanfamilies. Journal of Family Psychology, 14, 436–457.

Greene, M. L., Way, N., & Pahl, K. (2006). Trajectories of perceived adult and peer discriminationamong Black, Latino, and Asian American adolescents: Patterns and psychologicalcorrelates. Developmental Psychology, 42, 218–238.

Hill, N. E., Bush, K. R., & Roosa, M. W. (2003). Parenting and family socialization strategies andchildren’s mental health: Low-income, Mexican-American and Euro-American mothers andchildren. Child Development, 74, 189–204.

Holthouse, D. (2005). Arizona showdown: High-powered firearms, militia maneuvers and racism at theMinuteman project. Montgomery, AL: Southern Poverty Law Center.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structureanalysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6,1–55.

912 BERKEL ET AL.

Hughes, D., & Dodge, M. A. (1997). African American women in the workplace: Relationshipsbetween job conditions, racial bias at work, and perceived job quality. American Journal ofCommunity Psychology, 25, 581–599.

Huynh, V. W., & Fuligni, A. J. (2008). Ethnic socialization and the academic adjustment ofadolescents from Mexican, Chinese, and European backgrounds. Developmental Psychology,44, 1202–1208.

Kiang, L., Yip, T., Gonzales-Backen, M., Witkow, M., & Fuligni, A. J. (2006). Ethnic identity andthe daily psychological well-being of adolescents from Mexican and Chinese backgrounds.Child Development, 77, 1338–1350.

Klein, A., & Moosbrugger, H. (2000). Maximum likelihood estimation of latent interaction effectswith the LMS method. Psychometrika, 65, 457–474.

Klonoff, E. A., & Landrine, H. (1995). The schedule of sexist events: A measure of lifetimeand recent sexist discrimination in women’s lives. Psychology of Women Quarterly, 19,439–472.

Knight, G. P., Gonzales, N. A., Saenz, D., German, M., Deardorff, J., & Roosa, M. W. (2010). TheMexican American cultural values scale for adolescents and adults. Journal of Early Adoles-cence, 30, 444–481.

Landrine, H., & Klonoff, E. A. (1996). The schedule of racist events: A measure of racial dis-crimination and a study of its negative physical and mental health consequences. Journal ofBlack Psychology, 22, 144–168.

Lewinsohn, P. M., Zinbarg, R., Seeley, J. R., Lewinsohn, M., & Sack, W. H. (1997). Lifetimecomorbidity among anxiety disorders and between anxiety disorders and other mental dis-orders in adolescents. Journal of Anxiety Disorders, 11, 377–394.

Luthar, S. S., Cicchetti, D., & Becker, B. (2000). The construct of resilience: A critical evaluationand guidelines for future work. Child Development, 71, 543–562.

MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A com-parison of methods to test mediation and other intervening variable effects. PsychologicalMethods, 7, 83–104.

Manongdo, J. A., & Ramırez Garcıa, J. I. (2007). Mothers’ parenting dimensions and adolescentexternalizing and internalizing behaviors in a low-income, urban Mexican American sample.Journal of Clinical Child and Adolescent Psychology, 36, 593–604.

McMahon, S. D., Grant, K. E., Compas, B. E., Thurm, A. E., & Ey, S. (2003). Stress and psycho-pathology in children and adolescents: Is there evidence of specificity? Journal of Child Psy-chology and Psychiatry, 44, 107–133.

Midgley, C., Maehr, M. L., & Urdan, T. (1996). Patterns of adaptive learning survey (PALS). AnnArbor, MI: University of Michigan Press.

Murry, V. M., Berkel, C., Brody, G. H., Gerrard, M., & Gibbons, F. X. (2007). The Strong AfricanAmerican Families program: Longitudinal pathways to sexual risk reduction. Journal ofAdolescent Health, 41, 333–342.

Murry, V. M., Berkel, C., Brody, G. H., Miller, S. J., & Chen, Y.-F. (2009). Linking parentalsocialization to youth protective processes academic self-presentation and expectationsamong rural African American youth. Cultural Diversity and Ethnic Minority Psychology, 15,1–10.

Murry, V. M., Berkel, C., Pantin, H., & Prado, G. (2010). Family-based HIV prevention withAfrican American and Latino youth. In W. Pequegnat & C. C. Bell (Eds.), Families and HIV/AIDS. New York, NY: Springer Publishing. Manuscript submitted for publication.

Muthen, L., & Muthen, B. (2008). Mplus user’s guide (Version 5.1). Los Angeles, CA: Muthen &Muthen.

Pahl, K., & Way, N. (2006). Longitudinal trajectories of ethnic identity among urban Black andLatino adolescents. Child Development, 77, 1403–1415.

Perez, D. J., Fortuna, L., & Alegrıa, M. (2008). Prevalence and correlates of everyday discrim-ination among U.S. Latinos. Journal of Community Psychology, 36, 421–433.

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 913

Phinney, J. S., Madden, T., & Santos, L. J. (1998). Psychological variables as predictors of per-ceived ethnic discrimination among minority and immigrant adolescents. Journal of AppliedSocial Psychology, 28, 937–953.

Romero, A. J., Martinez, D., & Carvajal, S. C. (2007). Bicultural stress and adolescent riskbehaviors in a community sample of Latinos and non-Latino European Americans. Ethnicityand Health, 12, 443–463.

Romero, A. J., & Roberts, R. E. (1998). Perception of discrimination and ethnocultural variables ina diverse group of adolescents. Journal of Adolescence, 21, 641–656.

Romero, A. J., & Roberts, R. E. (2003a). The impact of multiple dimensions of ethnic identityon discrimination and adolescents’ self-esteem. Journal of Applied Social Psychology, 33,2288–2305.

Romero, A. J., & Roberts, R. E. (2003b). Stress within a bicultural context for adolescents ofMexican descent. Cultural Diversity and Ethnic Minority Psychology, 9, 171–184.

Romero, A. J., & Ruiz, M. (2007). Does familism lead to increased parental monitoring?Protective factors for coping with risky behaviors. Journal of Child and Family Studies, 16,143–154.

Roosa, M. W., Liu, F. F., Torres, M., Gonzales, N. A., Knight, G. P., & Saenz, D. (2008). Samplingand recruitment in studies of cultural influences on adjustment: A case study with MexicanAmericans. Journal of Family Psychology, 22, 293–302.

Roosa, M. W., Wolchik, S. A., & Sandler, I. N. (1997). Preventing the negative effects of commonstressors: Current status and future directions. In I. N. Sandler & S. A. Wolchik (Eds.), Hand-book of children’s coping: Linking theory and intervention (pp. 515–533). New York, NY: PlenumPress.

Ruggiero, K. M., Taylor, D. M., & Lambert, W. E. (1996). A model of heritage culture mainte-nance: The role of discrimination. International Journal of Intercultural Relations, 20,47–67.

Rumbaut, R. G. (1994). The crucible within: Ethnic identity, self-esteem, and segmentedassimilation among children of immigrants. International Migration Review, 28,748–794.

Shaffer, D., Fisher, P., Lucas, C. P., Dulcan, M. K., & Schwab-Stone, M. E. (2000). NIMH DiagnosticInterview Schedule for Children Version IV (NIMH DISC-IV): Description, differences fromprevious versions, and reliability of some common diagnoses. Journal of the American Academyof Child & Adolescent Psychiatry, 39, 28–38.

Scott, L. D. (2003). Cultural orientation and coping with perceived discrimination among AfricanAmerican youth. Journal of Black Psychology, 29, 235–256.

Shanahan, L., Copeland, W., Costello, E. J., & Angold, A. (2008). Specificity of putative psycho-social risk factors for psychiatric disorders in children and adolescents. Journal of Child Psy-chology and Psychiatry, 49, 34–42.

Smokowski, P. R., & Bacallao, M. L. (2006). Acculturation and aggression in Latino adolescents: Astructural model focusing on cultural risk factors and assets. Journal of Abnormal Child Psy-chology, 34, 659–673.

Smokowski, P. R., & Bacallao, M. L. (2007). Acculturation, internalizing mental health symptoms,and self-esteem: Cultural experiences of Latino adolescents in North Carolina. Child Psychi-atry and Human Development, 37, 273–292.

Stone, S., & Han, M. (2005). Perceived school environments, perceived discrimination, andschool performance among children of Mexican immigrants. Children and Youth Services Re-view, 27, 51–66.

Taylor, A. B., MacKinnon, D. P., & Tein, J.-Y. (2008). Tests of the three-path mediated effect.Organizational Research Methods, 11, 241–269.

Thayer, S. M., Valiente, C., Hageman, D., Delgado, M., & Updegraff, K. A. (2002). Conflict res-olution in friendships: A comparison of Mexican American and Anglo American adolescents. Paperpresented at the National Council of Family Relations, Houston, TX, USA.

914 BERKEL ET AL.

Umana-Taylor, A. J., Alfaro, E. C., Bamaca, M. Y., & Guimond, A. B. (2009). The central role offamilial ethnic socialization in Latino adolescents’ cultural orientation. Journal of Marriage andFamily, 71, 46–60.

Umana-Taylor, A. J., Bhanot, R., & Shin, N. (2006). Ethnic identity formation during adolescence:The critical role of families. Journal of Family Issues, 27, 390–414.

Umana-Taylor, A. J., & Updegraff, K. A. (2007). Latino adolescents’ mental health: Exploring theinterrelations among discrimination, ethnic identity, cultural orientation, self-esteem, anddepressive symptoms. Journal of Adolescence, 30, 549–567.

Unger, J. B., Ritt-Olson, A., Teran, L., Huang, T., Hoffman, B. R., & Palmer, P. (2002). Culturalvalues and substance use in a multiethnic sample of California adolescents. Addiction Researchand Theory, 10, 257–279.

U.S. Census Bureau. (2008). American Factfinder. Retrieved from http://factfinder.census.govWalls, M. L., Chapple, C. L., & Johnson, K. D. (2007). Strain, emotion, and suicide among Amer-

ican Indian youth. Deviant Behavior, 28, 219–246.Wayman, J. C. (2002). Student perceptions of teacher ethnic bias: A comparison of Mexican

American and non-Latino White dropouts and students. High School Journal, 85, 27–37.Wong, C. A., Eccles, J. S., & Sameroff, A. (2003). The influence of ethnic discrimination and ethnic

identification on African American adolescents’ school and socioemotional adjustment.Journal of Personality, 71, 1197–1232.

World Health Organization. (1993). The ICD–10 classification of mental and behavioural disorders.Diagnostic criteria for research. Geneva, Switzerland: Author.

DISCRIMINATION AND ADJUSTMENT FOR MEXICAN AMERICAN ADOLESCENTS 915