Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Jul 18.
Published in final edited form as: J Early Adolesc. 2009 Feb;29(1):16–-42. doi: 10.1177/0272431608324475

Familism Values as a Protective Factor for Mexican-origin Adolescents Exposed to Deviant Peers

Miguelina Germán 1, Nancy A Gonzales 1, Larry Dumka 1
PMCID: PMC3138713  NIHMSID: NIHMS114595  PMID: 21776180

Abstract

This study examined interactive relations between adolescent, maternal and paternal familism values and deviant peer affiliations in predicting adolescent externalizing problems within low-income, Mexican-origin families (N = 598). Adolescent, maternal and paternal familism values interacted protectively with deviant peer affiliations to predict lower levels of externalizing problems according to two independent teacher reports. These relations were not found with parent reports of adolescent externalizing problems although these models showed a direct, protective effect of maternal familism values. Consistent with the view that traditional cultural values are protective for Latino adolescents, these results suggest that supporting familism values among Mexican-origin groups is a useful avenue for improving adolescent conduct problems, particularly in a school context.

Keywords: familism values, deviant peers, externalizing, low-income families, Mexican origin, protective factor

Exposure to deviant peers puts Mexican-origin teens at risk for substance use, antisocial behavior, and externalizing problems (Barrera, Biglan, Ary, & Li, 2001; Brooks, Stuewig, & Lecroy, 1998). Previous research indicates that deviant peer activity tends to be higher in low-income communities (Tolan, Guerra, & Montaini-Klovdahl, 1997); thus, Mexican-origin adolescents may have more delinquent peer exposure in comparison to other groups because they are overrepresented among America’s poor. Mexican-origin people have twice the rate of poverty of other groups in the U.S. and comprise 60% of Latinos, the largest ethnic minority group in the nation (Camarota, 2001). Vigil (1997) writes that adolescents gangs, a breeding ground of delinquent activity, have become a fixture among Mexican-American neighborhoods and that as many as 10% of Mexican-origin adolescents belong to gangs.

In light of these trends, identifying protective factors for Mexican-origin adolescents living in low-income, high risk peer environments is critical to inform culturally sensitive treatment and prevention. This study suggests that traditional Latino family values, often referred to as familism, may function as a protective factor by mitigating the negative effects of deviant peer exposure on externalizing problems. Examining familism values as a protective resource is consistent with recent recommendations in child development to study adaptive aspects of culture (Garcia Coll, Akerman, & Cicchetti, 2000). Morevoer, Cauce and Domenech-Rodriguez (2002) specifically urged Latino researchers to move beyond simple main effect designs and study the conditions under which core cultural values such as familism represent a source of strength or weakness for Latino families.

Familism Values as a Protective Factor

This study defines familism values as a set of normative beliefs espoused by Latino populations that emphasize the centrality of the family unit and stress the obligations and support that family members owe to both nuclear and extended kin (Sabogal, Marin, Otero-Sabogal, Marin, & Perez-Stable, 1987). Sabogal and colleagues (1987) conducted a factor analytic study that revealed three related facets of familism values. The first facet, familial obligations, is defined as the belief that family members have a responsibility to provide economic and emotional support to kin. The second facet, perceived support and emotional closeness, is defined as the perception that family members are dependable sources of help, should be united, and have close relationships. The third facet, family as referent, is the belief that family members’ behaviors should meet with familial expectations. Within this referential framework the family is viewed as an extension of oneself such that the behavior of an individual is a reflection of the whole family. Given the theoretical and empirical interrelatedness of these subscales, this study used a measure that incorporates all three facets into one overall measure of familism values (Knight, Gonzales, Saenz, Germán, Deardorff, Roosa, & Updegraff, under review).

Familism values are considered core cultural values for Latinos that are transmitted from generation to generation through socialization strategies and interactions that parents have with their children (Bronfenbrenner, 1979). Roosa and colleagues (2002) theorize that familism values guide Latino parents’ selection of childrearing strategies that directly promote youth behavior that is consistent with these values. For example, ethnographies and survey studies have highlighted Mexican cultural traditions and socialization practices that emphasize the role of children in supporting, assisting, and respecting the authority of the family (Updegraff, McHale, Whiteman, Thayer, & Delgado, 2005; Valdés, 1996). In the case of immigrant families, it is typical for bilingual children to help their Spanish monolingual parents and extended kin navigate English-dominant systems such as schools, hospitals and financial institutions. Children’s behavior in public is viewed as a reflection on the family unit and thus, parents frequently direct children to behave appropriately in public settings and expect older siblings to be responsible for and monitor younger siblings.

The notion that these value dimensions operate to protect adolescents against delinquent peer activity is consistent with the theory of social control, which emphasizes the importance of adolescent bonds to conventional institutions such as the family unit in preventing delinquent behavior (Hirschi, 1969). It is expected that familism values operate to cement strong bonds of attachment to the family unit and ensure that the family continues to be a strong source of support and guidance even as developing adolescents become increasingly involved with peers and activities outside the home (Keefe, Padilla, & Carlos, 1978). Internalization of familism values, in turn, may help adolescents exert self-control to not model their peers’ deviant behaviors. To the extent that an adolescent has close ties to the family and believes her behavior will reflect on the family unit, she may be less likely to conform with delinquent peers out of a fear of disappointing significant others (Vega, Gil, Warheit, Zimmerman, & Apospori, 1993).

A few studies have found direct, protective effects of familism values; these studies report negative associations between adolescent familism values and deviant behavior (Gil, Warheit, Zimmerman & Apospori, 1993), interpersonal violence (Sommers, Fagan, & Baskin, 1993) and cigarette use (Kaplan, Napoles-Springer, Stewart, & Perez-Stable, 2001). However, we know of only one study that examined whether familism values operate interactively to diminish the influence of deviant peers on adolescent problem behavior. Brook, Whiteman, Balka, Win, and Gursen (1998) found that the negative effects of peer illegal drug use on adolescent stage of drug use were diminished for Puerto Rican adolescents who reported higher levels of familism values. Given that deviant peer affiliation is such a robust predictor of problem behavior for adolescents, it is important to explore whether familism values function to moderate the impact of deviant peers on problem behavior more generally, not just with respect to illegal drug use. This study contributes to the current research by examining familism values as a protective factor for the negative effects of deviant peer affiliation on externalizing problems in adolescence.

Another limitation of the literature is that, to date, most of the empirical work on familism values and deviant peers has focused on the adolescents’ familism values. Despite theoretical work to suggest that parental familism values may also function to protect adolescents exposed to deviant peers, no studies to date have tested the impact of maternal and paternal familism values on adolescent behavior. Similar to the effects of adolescent familism values suggested above, parental values emphasizing the importance of emotional closeness, family as referent, and spending time together within the family also are likely to strengthen parent-child bonds of attachment, promote greater family cohesion, and motivate adolescents to adhere to the family’s prosocial norms (Oetting & Donnermeyer, 1998). Thus, this study also examined the putative protective effects of adolescent, maternal, and paternal familism values.

Goals of the Present Study

The primary goal of the current study was to examine the interactive effects of familism values and deviant peer affiliations as predictors of adolescent externalizing problems. This study uses a sample of Mexican-origin, early adolescents following the transition to large, urban middle schools in low-income, high-crime neighborhoods. This is a critical context and developmental period to examine the protective benefits of familism values because conformity to deviant peers peaks and adolescent engagement in school decreases during this period (Seidman, Allen, Aber, Mitchell, & Feinman, 1994; Vitaro, Tremblay, Kerr, Pagani, & Bukowski, 1997), particularly for minority adolescents (Seidman, Lambert, & Allen, 2001).

To control for adolescents’ social desirability response bias in reporting their own externalizing problems, this study used four adult reports of adolescent externalizing problems – mother, father and two teacher reports. Having multiple raters allowed an assessment of the influence of shared method variance and an assessment of the consistency of the results across different raters of adolescent outcome. This multiple reporter assessment strategy allowed us to examine the following questions: (1) Do adolescent, maternal and paternal familism values provide interactive protective effects to reduce deviant peer influences on adolescent externalizing behavior? (2) Do the interactive protective effects of familism values operate similarly across maternal, paternal and teacher reports of adolescent behavior?

Method

Participants

Five hundred and ninety-eight 7th grade students and their parents were recruited as part of an intervention study from five junior high schools in the metropolitan Phoenix area that serve primarily low-income, Mexican-origin students; 80% of students at these schools were enrolled in the state’s free lunch program and 82% identified as Hispanic. All the students had just experienced a school transition from elementary school to junior high school and lived in school districts with some of the highest school dropout rates and crime statistics in the Phoenix metropolitan area.

We recruited English- and Spanish-speaking Mexican-origin families into a randomized trial of the Bridges to High School Program, an intervention study designed to prevent high school dropout and decrease the incidence of mental and behavioral health disorders in adolescents. Participating schools provided access to school rosters from which students were randomly selected for recruitment. Of those selected, 27% were unable to be located using the school-provided information and 8% refused to be screened. To be eligible, the student and at least one parent had to identify as Mexican origin and agree to be randomized to an intervention study that would either place them in a one-session control group or an 11-week intervention group. In addition, the project required that both parents and adolescents be able to participate in the project in the same language. Of the families determined eligible, 75% agreed to accept these conditions and enroll in the study. The current investigation uses pretest data collected from these families prior to randomization to intervention condition.

Of the 598 students, 303 (50.6%) were female, and 295 (49.2%) were male. The adolescents ranged in age from 11 to 14 years, with a mean age of 12.3 years. Three hundred and nineteen adolescents (53.4%) were interviewed in Spanish and 278 in English (46.6%). Four hundred and ninety-five adolescents (82.8%) lived in two-parent families and 103 (17.2%) lived in one-parent families. The mean household per capita income was $7,220 with a standard deviation of $5,092.

Of the parents, 573 mothers and 331 fathers participated in the interviews. Among the mothers, 314 (54.8%) were interviewed in Spanish and 259 (45.2%) were interviewed in English. Maternal education levels (M = 9.89, SD = 3.56, range = 0 to 18) are as follows: 339 (59.3%) did not graduate high school, 96 (16.8%) were high school graduates, 75 (13.1%) had some college or vocational experience, 56 (9.8%) held vocational, associate or college degrees, and 6 (1.0%) had some post-graduate work or advanced degrees. Among the fathers, 200 (60.4%) were interviewed in Spanish and 131 (39.6%) were interviewed in English. Paternal education levels (M = 10.01, SD = 3.69, range 0 to 20) are as follows: 185 (56.1%) did not graduate high school, 64 (19.4%) were high school graduates, 49 (14.8%) had some college or vocational experience, 27 (8.1%) held vocational, associate or college degrees, and 6 (1.5%) had some post-graduate work or advanced degrees.

Procedure

For three consecutive years, the study randomly selected eligible participants from 7th grade school rosters and recruited families via mail and telephone. In-home interviews were scheduled and conducted by trained interviewers using laptop computers. Interviewers were trained to conduct the parent and target child surveys in separate rooms and/or out of the hearing of other family members. The interviewers read each survey question and possible responses aloud in either Spanish or English to reduce problems associated with variations in literacy levels. Each member of the family who completed an interview received $30, for a total of $60 for one-parent and $90 for two-parent families. Families were asked to identify a language arts teacher and a math teacher (hereafter described as teacher #1 and #2, respectively) currently instructing the adolescent and gave permission for us to contact them. Teachers were mailed questionnaires regarding student behavior and paid $5 for each completed survey. Response rates were over 90% for both sets of teachers.

Measures

Deviant Peer Affiliations

The study employed a 15-item scale to assess the degree to which adolescents associated with deviant peers. Barrera et al. (2001) developed this scale from several scales previously used in research with adolescents (Dishion, Patterson, Stoolmiller, & Skinner, 1991; Mason, Cauce, Gonzales, & Hiraga, 1995). Adolescents were asked to indicate how many of their peers had engaged in particular deviant or antisocial activities (e.g., “gotten drunk or high”, “started a fight with someone”) during the past month with responses ranging from 1 (none) to 5 (almost all). The measure had a coefficient alpha equal to .90.

Familism values

The 16-item Familism Values Scale was taken from three subscales of a larger measure of traditional cultural and mainstream values for Mexican-origin adolescents and adults (Knight et al., under review). Items for this measure were derived from focus group interviews in which Mexican-origin mothers, fathers, and adolescents were asked to describe values and attitudes that reflected their traditional Mexican culture and values they perceived as important within mainstream U.S. culture. Focus groups were conducted in English and Spanish and in diverse community contexts to ensure variability on participants’ values. The focus group data identified aspects of familism that are similar to those identified in other familism scales in the literature (Bardis, 1959; Heller, 1970; Sabogal et al., 1987) and provided qualitative descriptions to operationalize these dimensions with culturally grounded items.

The familism subscales are composed of three dimensions that assessed values toward the family unit and its members including beliefs regarding obligations to the family, appropriate levels of emotional closeness and support, and using the family as a referent when making decisions. Parents and adolescents were asked to rate how much they agreed or disagreed with each item with responses ranging from 1 (strongly disagree) to 5 (strongly agree). Although stability coefficients are not yet available for the familism subscales, Knight et al. (under review) established their validity using nativity (country of birth), the Acculturation Rating Scale for Mexican Americans-II (ARSMA-II; Cuellar, Arnold, & Maldonado, 1995), and a measure of Mexican American ethnic pride (MAEP: Gonzales, Knight, & Saenz, 2005). Given that the familism subscales are thought to capture traditional Mexican values, these subscales were expected to correlate positively with being born in Mexico, the Mexican Orientation scale from the ARSMA-II and the ethnic pride scale. As hypothesized, this pattern of relations emerged. The pearson product moment correlations among the subscales ranged from .50 to .64 across reporters in the current sample. Consequently, the subscales were combined to form an overall construct of familism values. Coefficient alpha equaled .84 for adolescent, .80 for maternal, and .77 for paternal reports.

Externalizing problems

Externalizing problems were assessed by maternal and paternal reports on the Child Behavior Checklist (CBCL) and two separate teachers on the Teacher Report Form (TRF; Achenbach, 2001). The externalizing grouping is defined as the sum of the scores on the Delinquent and Aggressive Behavior subscales. Each item is answered using a 3-point Likert-type scale (e.g., “Not true” to “Very true or often true”). The percentage of adolescents in the current sample that fell within the clinical range on the externalizing scale was 11.6% for maternal report, 8.7% for paternal report, 12.6% for teacher #1, and 8.7% for teacher #2. These percentages were either comparable or higher than the percentages of the normative sample’s scores (clinical = 10%; Achenbach, 2001), although it should be noted that the normative sample excluded non-English speaking parents. Coefficient alpha was .89 for maternal report, .91 for paternal report, .93 for teacher #1, and .94 for teacher #2.

Socioeconomic Status (SES)

Socioeconomic status was represented as a composite of parental education level and occupation. SES was created using the highest level of each indicator achieved by the male or female caregiver in the family. The composite score is the mean of the standardized z-scores of these two measures which were correlated, r = .28, p < .05.

Results

The intercorrelations, means, and standard deviations for all study variables, including adolescent gender, SES, predictors, and outcome variables are presented in Table I. It is notable that maternal and paternal reports of adolescent externalizing problems were more highly significantly correlated with each other than with either of the teacher ratings of adolescent externalizing problems. The same pattern was observed between the two teacher ratings. R to Z transformations were conducted with 95% confidence intervals, and in all cases these differences were statistically significant (Cohen, Cohen, West, & Aiken, 2003). It is also surprising that youth familism values were not correlated with either maternal or paternal familism values.

Table I.

Intercorrelations & Descriptive Statistics

Variable 1 2 3 4 5 6 7 8 9 10
1. Deviant Peer Affiliation -- −.08 −.07 .04 .27** .27** .34** .33** .001 .06
2. Adolescent Familism Values -- .04 .03 −.05 −.01 −.07 −.11** .02 .02
3. Maternal Familism Values -- .18** −.14* −.17** −.08 −.06 −.02 −.16**
4. Paternal Familism Values -- −.07 .01 −.18** −.08 .05 −.17**
5. Maternal Externalizing Report -- .53** .34** .27** −.05 −.05
6. Paternal Externalizing Report -- .34** .34** −.03 −.07
7. Teacher #1 Externalizing Report -- .54** −.17** −.01
8. Teacher #2 Externalizing Report -- −.09* .003
9. Adolescent Gender -- −.12*
10. SES --
n 597 596 573 331 573 330 539 555 598 594
M 1.54 4.46 4.45 4.55 8.17 7.41 4.67 3.50 1.51 7,220
SD .55 .42 .42 .34 7.24 7.31 7.75 6.23 .50 5,092
Open in a new tab
*

p < .05.

**

p < .01.

Regression Models

A series of 12 hierarchical regression analyses tested the hypothesized relationships. For each of the 4 reports of externalizing problems (maternal, paternal, teacher #1, and teacher #2), a separate analysis was conducted to test for the protective effects of adolescent, maternal and paternal familism values. Each regression analysis included familism values, deviant peer affiliation, and the interaction between familism values and deviant peer affiliation as predictors. Gender and SES were considered as potential covariates as extant research suggests these variables have been consistently associated with adolescent externalizing problems in previous research (Crick & Zahn-Waxler, 2003; Barrera et al., 2002). While gender demonstrated bivariate relations with adolescent outcome in this sample, SES did not and thus was dropped as a predictor from the regression models presented. Preliminary analyses were first conducted to test all covariate by predictor interactions (e.g., gender × familism values). Some significant two-way interactions between the predictors and gender were found in a few regression models, and these terms were included in those specific models only.

Centering, Coding and Regression Diagnostics

All predictor variables were centered to eliminate nonessential multicollinearity and to facilitate the interpretation of lower order coefficients in the presence of an interaction (Aiken & West, 1991). Moreover, adolescent gender was coded using weighted effects coding to accurately reflect the proportion of boys and girls in the sample which yielded a centered dichotomous predictor (West, Aiken, & Krull, 1996). Regression diagnostics were examined to detect data points with disproportionate influence and to determine the extent to which these cases might affect the regression outcomes. These analyses revealed no problems with multicollinearity or undue influence.

Analyses with Adolescent Familism values

Two significant interactions between adolescent familism values and deviant peer affiliation were found with teachers #1 and 2 reports of externalizing behaviors but not with maternal or paternal reports (see Table II). As expected, higher levels of deviant peers consistently predicted more externalizing problems, regardless of reporter. It is interesting to note that adolescent familism values had no significant direct effects in any of these models. The adolescent familism models accounted for 11 to 14% of the variance in externalizing behaviors as reported by teachers while the parent report models accounted for 5 to 7% of the variance.

Table II.

Hierarchical Regression Analyses Examining Adolescent Familism Values, Deviant Peer Affiliation, and Gender as Predictors of Adolescent Externalizing Problems (as reported by Mother, Father, Teacher #1, and Teacher #2)

Variable Mother Report of Externalizing Father Report of Externalizing Teacher #1 Report of Externalizing Teacher #2 Report of Externalizing
 AFam .01 −.03 −.02 −.07
 DPeer .22*** .25*** .30*** .28***
 Gend −.04 −.07 −.18*** −.11**
 Afam × DPeer −.004 .07 −.12** −.07*
Total R2 .05 .07 .14 .11
Total F F(4, 566) = 7.31*** F(4, 324) = 5.74*** F(4, 533) = 21.37*** F(4, 549) = 16.72***
Open in a new tab

Note. Afam = adolescent familism values; DPeer = deviant peer affiliation; Gend = youth gender. βs are standardized regression

coefficients. All data are from the final step of the models.

*

p < .05;

**

p < .01;

***

p < .001.

The interaction between adolescent familism values and deviant peers was probed as a function of the familism variable (one standard deviation above the mean versus one standard deviation below the mean) using simple slope analyses as recommended by Aiken and West (1991). The analysis for teacher #1 report of externalizing problems indicated that the relation between deviant peers and externalizing problems was stronger at lower levels of adolescent familism values, β = .41, p < .001, compared with higher levels of familism values, β = .17, p < .01. The analysis for teacher #2 report of adolescent outcome showed a similar pattern between deviant peers and externalizing problems for youth at one standard deviation below the mean on familism values, β = .35, p < .001 compared with one standard deviation above the mean on familism values, β = .20, p < .01. For illustrative purposes, Figure 1 presents the simple slopes for the relation of deviant peer affiliation to adolescent externalizing problems for the teacher #1 outcome model. Since the other interaction involving adolescent familism values was of the same form, it was not plotted.

Figure 1.

Figure 1

Open in a new tab

Plot of the simple slopes for the relation between deviant peer affiliations and adolescent externalizing problems at low familism values (−1 standard deviation) and high familism values (+1 standard deviation).

Analyses with Maternal Familism values

One significant interaction was found between maternal familism values and deviant peer affiliation with teacher #2 report of adolescent outcome but not in the outcome models as reported by teacher #1, mothers or fathers. (see Table III). However, the parent outcome models did have significant main effects of maternal familism values such that higher values were associated with lower levels of adolescent externalizing problems. Higher levels of deviant peer affiliations were associated with higher levels of adolescent externalizing problems in the parental reports of outcome but interacted with gender in the teacher report models. The maternal familism models with teacher report of externalizing behaviors accounted for 11 to 14% of the variance while the parent report models accounted for 8 to 9% of the variance.

Table III.

Hierarchical Regression Analyses Examining Maternal Familism Values, Deviant Peer Affiliation, and Gender as Predictors of Adolescent Externalizing Problems (as reported by Mother, Father, Teacher #1, and Teacher #2)

Variable Mother Report of Externalizing Father Report of Externalizing Teacher #1 Report of Externalizing Teacher #2 Report of Externalizing
 MFam −.16*** −.14* −.06 −.04
 DPeer .21*** .24*** .31*** .29***
 Gend −.05 −.09 −.19*** −.12***
 Mfam × DPeer .02 −.02 −.05 −.09*
 Dpeer × Gend n/a n/a −.10* −.11*
Total R2 .08 .09 .14 .11
Total F F(4, 567) = 11.63*** F(4, 301) = 7.04*** F(5, 508) = 15.98*** F(5, 523) = 13.39***
Open in a new tab

Note. Mfam = maternal familism values; DPeer = deviant peer affiliation; Gend = youth gender. βs are standardized regression coefficients.

All data are from the final step of the models. n/a Predictor not included in the model.

*

p < .05;

**

p < .01;

***

p < .001.

The plot of the interaction between maternal familism values and deviant peers in the model reported by teacher #2 had the same form as the interaction graphed in Figure 1. Simple slope analyses indicated a weaker relation between deviant peer affiliations and externalizing problems for youth who had mothers endorsing familism values at one standard deviation above the mean (β = .20, p < .01.), compared with mothers with familism values at one standard deviation below the mean (β = .33, p < .001.). The two-way interactions of deviant peer affiliation and adolescent gender in the teacher report models revealed that for boys, there was a stronger relationship between deviant peer affiliation and externalizing problems than for girls. This effect was found in the model predicted by teacher #1 for boys (β = .41, p < .001) versus girls (β = .24, p < .001) and in the model predicted by teacher #2 for boys (β = .39, p < .001) versus girls (β = .25, p < .001).

Analyses with Paternal Familism values

Two significant interactions between paternal familism values and deviant peer affiliation were found with teachers #1 and 2 reports of externalizing behaviors but not with maternal or paternal reports (see Table IV). Moreover, additional interactive effects emerged in the teacher report models. An interaction between gender and paternal familism values emerged in teacher #1 model and an interaction between deviant peer affiliation and gender emerged in the model with teacher #2 report of outcome. Across reporters, higher levels of deviant peers consistently predicted more externalizing problems. Only the teacher #1 outcome model had a direct, protective effect of paternal familism values. The paternal familism models accounted for 13 to 21% of the variance in externalizing behaviors as reported by teachers while the parent report models accounted for 4 to 7% of the variance.

Table IV.

Hierarchical Regression Analyses Examining Paternal Familism Values, Deviant Peer Affiliation, and Gender as Predictors of Adolescent Externalizing Problems (as reported by Mother, Father, Teacher #1, and Teacher #2)

Variable Mother Report of Externalizing Father Report of Externalizing Teacher #1 Report of Externalizing Teacher #2 Report of Externalizing
 PFam .002 −.08 −.18** −.05
 DPeer .20** .24*** .27*** .22***
 Gend −.06 −.06 −.21*** −.15**
 Dpeer × Pfam .03 −.05 −.12* −.12*
 Pfam × Gend n/a n/a .14** −.05
 Dpeer × Gend n/a n/a n/a −.15*
Total R2 .04 .07 .21 .13
Total F F(4, 301) = 3.12* F(4, 325) = 6.20*** F(5, 295) = 15.63*** F(6, 312) = 7.99***
Open in a new tab

Note. Pfam = paternal familism values; DPeer = deviant peer affiliation; Gend = youth gender. βs are standardized regression coefficients. All data

are from the final step of the models. n/a Predictor not included in the model.

*

p < .05;

**

p < .01;

***

p < .001.

Probing the interaction between paternal familism values and deviant peers indicated that youth with fathers at one standard deviation above the mean on familism values showed no relation between deviant peer affiliation and externalizing problems as reported by teacher #1 (β = .12, ns) or teacher #2 (β = .03, ns). In contrast, the simple slope analyses for fathers with lower familism values revealed a significant relation between deviant peers and externalizing problems as reported by teacher #1, (β = .43, p < .001) and teacher #2 (β = .38, p < .001). The plots of these interactions were similar in form to Figure 1.

The two-way interaction of paternal familism values and adolescent gender in the regression model reported by teacher #1 revealed that the association between fathers’ familism values and externalizing problems was significant for boys (β = −.39, p < .001) but not for girls (β = .04, p = .639). More specifically, boys with fathers who reported higher levels of familism values exhibited less externalizing problems while there was no relation between fathers’ level of familism values and externalizing problems for girls. One other conditional effect of gender was found with deviant peer affiliation in the regression model predicted by teacher #2 such that there was a stronger relation between deviant peer affiliation and externalizing problems for boys (β = .44, p < .001) versus girls (β = .05, p = .495).

Discussion

The main goal of this investigation was to examine the interactive protective effects of familism values and deviant peer affiliation on adolescent externalizing problems. This question was addressed in a sample of low-income, Mexican-origin adolescents, all of whom recently transitioned to junior high schools located in two inner-city school districts. Greater exposure to deviant peers was consistently associated with increased youth externalizing problems, supporting an extensive literature documenting relations between deviant peer affiliations and youth conduct problems (Ary et al., 1999; Barrera, et al., 2001; Barrera, et al., 2002). However, the negative impact of deviant peer affiliations on adolescent externalizing behavior was attenuated by adolescent, maternal and paternal familism values in five of six models that were tested with teacher reports of externalizing behaviors. These findings support our central hypothesis that traditional cultural values are protective for Mexican origin youth (Samaniego & Gonzales, 1999). On the other hand, the interaction between familism values and deviant peer affiliation was not significant using maternal and paternal reports of adolescent externalizing behaviors. The following discussion focuses on findings that emerged in the teacher report models, followed by a discussion of possible reasons these effects were not replicated in the parent report models.

Familism Effects on Teacher Reports of Externalizing

Results of this study revealed that adolescent, maternal, and paternal familism values function interactively with deviant peer affiliation to protect adolescents from higher levels of externalizing problems as reported by teachers. Replication of the interactive protective effects of familism across two independent teacher reports increases confidence in these results, particularly given that these findings are not biased by method variance. These findings are consistent with a study that reported adolescent familism values mitigated the negative effects of peer drug use on adolescents’ stage of drug use (Brook et al., 1998) and extends these results to a broader measure of deviant and aggressive behaviors. In addition, this study demonstrates that adolescent and parental familism values may make Mexican-origin adolescents less vulnerable to the negative effects of deviant peers.

It is interesting to note that the strongest interactive protective effects were found for paternal familism values. In fact, the typically robust relation between deviant peer affiliation and teacher report of externalizing problems was reduced to non-significance when fathers endorsed high levels of familism. Although research on Mexican-origin fathering and its influences on child development is scarce, emerging evidence is beginning to show that Latino fathers may be more involved in monitoring and supervising their children than Anglo fathers (Toth & Xu, 1999) and that Mexican-origin fathers play an important, positive role in shaping their children’s behavior through their provision of parental support, discipline and monitoring (Formoso, Gonzales, Barrera, & Dumka, in press; Parke, 2004). Although not directly tested in the current study, one critical role that traditional familism values may play within Mexican-origin families is to keep fathers involved in these positive ways that may be especially important to offset the negative influence of deviant peers in low-income communities.

Significant direct and interactive effects of adolescent gender were also shown in the teacher report models. As expected, teachers reported higher levels of externalizing for boys than for girls, confirming expected gender differences (Crick & Zahn-Waxler, 2003). In addition, a significant interaction between gender and deviant peer affiliation indicated that boys were more likely than girls to show an increase in externalizing behaviors as their exposure to deviant peers increased; this interactive effect only emerged in the teacher report models as well. Finally, a significant interaction between paternal familism values and gender emerged in one of the two teacher report models, showing that familism has a stronger relation to externalizing problems for boys than for girls. Together, these findings highlight the greater vulnerability of adolescent males within a deviant peer context but also suggest that supporting the role of Mexican-origin fathers may be a useful avenue to reduce this heightened vulnerability.

Familism Effects on Parent Report of Externalizing: Lack of Replication

The lack of significant interactive effects in models with parent report of externalizing raises the possibility that the buffering effects of familism are specific to the school context or to teachers’ perceptions of problem behavior. Achenbach et al. (1987) suggested that parents and teachers “are not exposed to identical samples of a child’s behavior, both because they see the child in different contexts and because they interact differently with the child” (p. 214). In support, we found higher concordance in ratings of externalizing behaviors reported by the two teachers and the two parents than were found across these two types of reporters. Assuming all reports are equally valid, one possible interpretation of the current pattern of findings is that familism values may function differently with respect to externalizing behaviors in the school versus home contexts.

For example, because familism values emphasize the belief that an individual’s behavior is a reflection on the family, it is possible they exert a stronger constraining influence on adolescents’ behavior in public settings, particularly in situations when their behavior is likely to be evaluated by authority figures (i.e., teachers). Qualitative research has shown more traditional Mexican-origin families, particularly immigrants, place high value on raising a child who is well mannered and respectful of authority figures (Valdés, 1996). Sirolli (2004) found more traditional, Spanish-speaking parents are more critical when children behave poorly in public and are more likely than more acculturated parents to blame child misbehavior on the family unit. Valdés (1996) also described Mexican-origin children who were normally very boisterous and energetic at home behaving in a controlled and contained manner at school. Thus, well-mannered behavior (“bien educado”) outside the family context may be an important way that familistic values are demonstrated within Mexican origin families.

Research also has shown that a strong sense of family obligation is associated with greater academic motivation, particularly the belief in the importance and usefulness of education (Fuligni, Tseng, & Lam, 1999; Fuligni, 2001). Fuligni (2001) and Suárez-Orozco and Suárez-Orozco (1995) suggest that this aspect of familism provides extra motivation to do well in school because it is associated with a desire to support and assist the family. Thus, it is possible that familism operates to mitigate behavior in the school context to a greater extent because it is uniquely associated with adolescents’ motivation to stay out of trouble and succeed in school.

Alternatively, it is possible that teachers and parents utilize different standards of judgment when rating externalizing problems (Achenbach et al., 1987). Whereas teachers’ frame of reference for normal and abnormal behavior is influenced by the fact that they interact with hundreds of students, parents may have more limited exposure to adolescent behavior. Perhaps their reports failed to capture variations in problem behavior as accurately as teachers for this sample. In fact, there was no relation of adolescent gender to parent reports of externalizing problems, which is contradictory to prior research. Thus, given the paucity of studies that have examined the interactive protective effects of familism values on problem behavior outcomes with multiple reporters, it would be premature to draw strong conclusions based on our study’s null findings using parental reports of adolescent externalizing problems. However, it is important to note that if this study had not examined both the direct and interactive effects of familism values with multiple reports of adolescent outcome, it might have led to the erroneous conclusion that familism values were unrelated to adolescent externalizing problems.

It also is important to note that, although the interaction between familism and deviant peer affiliation was not significant using parent reports of adolescent outcome, maternal familism had a significant main effect on maternal and paternal reports of externalizing. Mexican-origin mothers have been described as the primary socialization agent responsible for maintaining family values and structuring the family environment to reflect these values (Valdés, 1996). To the extent that mothers hold strong familism values, adolescents may be more likely to exhibit prosocial, respectful behavior when interacting with parents and other family members in the home context, regardless of their exposure to deviant peers.

Study Limitations and Implications

Although this study offers a unique contribution to psychological research with Mexican-origin families, several limitations should be noted. First, although we theorized that parents socialize their children to have similar values, there was a lack of correlation between youth and parents’ familism values. Children of immigrants are exposed to different sets of cultural values – both mainstream American values and their parents’ traditional values from their country of origin (Phinney, Ong, & Madden, 2000). During adolescence, peer influences rise and it becomes normative to engage in identity exploration. It may be that for immigrant families, intergenerational discrepancies in cultural values may increase between parents and adolescents if the original heritage and the new culture do not endorse identical values. This may potentially explain the lack of correlation between youth and parental familism values found in this study.

Another limitation of this study is that we did not control for immigrant status or language of interview, two proxy variables typically used to measure acculturation level. Several studies have reported an association between increased acculturation and adolescent problem behaviors among Latino groups (Gonzales, Knight, Morgan-Lopez, Saenz, & Sirolli, 2002). Although a loss of traditional family values are often offered as an explanation for this association (Buriel et al., 1982; Samaniego & Gonzales, 1999; Sommers et al., 1993), prior studies have not tested the relation between specific cultural values and adolescent problem behavior that might account for this link. In this sample, immigrant status and language of interview were not appropriate to use as covariates because they were more strongly correlated with familism values than with adolescent externalizing behaviors. Although familism values may be important in understanding the association between the acculturation process and externalizing problems, the covariation shared among familism values and other markers of acculturation make the interrelations among these variables difficult to parse apart in a cross-sectional design. Alternatively, the relations among these different markers of the acculturation process may be better modeled within a mediational framework.

This study’s cross-sectional design only allows for an assessment of concurrent relations between variables; thus, the model is subject to several alternative explanations. It is possible that Mexican-American youth with lower or higher rates of externalizing problems impact the familism values of their parents and that the relations among these variables are not linear but recursive and transactional. Future research should examine the relations among familism values and diverse youth outcomes in a longitudinal design to assess direct, indirect and conditional relations.

Despite these limitations, the current study examined a construct that is emic, or culturally specific to a particular group and, in so doing, identified a potentially important cultural resource. This is an important contribution to the current literature because it allows for the identification of protective factors that may be overlooked if studies on child development only examine factors and processes that are universal in nature. For persons that aim to promote positive outcomes among Mexican-American youth, a better understanding of culturally specific risk and protective factors is vital to increase cultural competence (Dumka, Lopez, & Carter, 2002).

The study’s robust finding that familism values have protective effects on adolescent behavior in the school context is noteworthy because of the disproportionately higher rates of school problems and academic failure reported for Mexican origin youth in the U.S. (Chavez, Oetting, & Swaim, 1994; U.S. Department of Education, 2000). If the results of this study are replicated, it may carry critical policy implications because some school policies view the ethnic culture of Latino families, particularly immigrant families, as an impediment to school success. Gonzales, Knight, Birman, and Sirolli (2004) argue that school policies that build on existing strengths in traditional Latino culture and favor cultural competence training among school personnel would better promote adaptation and success in school for Latino youth. A better understanding of the protective benefits of familism values for Mexican-origin populations and other Latino groups has the potential to improve behavioral and academic outcomes for these youth.

Acknowledgments

The research was supported by NIMH grant 1-R01-MH64707-01 to fund a Preventive Intervention for Mexican American Adolescents. The authors acknowledge Mark Roosa, George Knight, and Lorey Wheeler for their important contributions to this research.

Footnotes

This article may not exactly replicate the final version published in the journal. It is not the copy of record.

References

  1. Achenbach TM, McConaughy TM, Howell CT. Child/adolescent behavioral and emotional problems: Implications of cross-informant correlations for situational specificity. Psychological Bulletin. 1987;101:213–232. [PubMed] [Google Scholar]
  2. Achenbach TM, Rescorla LA. Manual for the ASEBA school-age forms & profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families; 2001. [Google Scholar]
  3. Aiken L, West S. Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage; 1991. [Google Scholar]
  4. Ary DV, Duncan TE, Biglan A, Metzler CW, Noell JW, Smolkowski K. Development of Adolescent Problem Behavior. Journal of Abnormal Child Psychology. 1999;27:141–150. doi: 10.1023/a:1021963531607. [DOI] [PubMed] [Google Scholar]
  5. Bardis PD. Attitudes toward the family among college students and their parents. Sociology and Social Research. 1959;43:352–358. [Google Scholar]
  6. Barrera M, Jr, Biglan A, Ary D, Li F. Replication of a Problem Behavior Model with American Indian, Hispanic, and Caucasian Adolescents. Journal of Early Adolescence. 2001;21:133–157. [Google Scholar]
  7. Barrera M, Jr, Prelow HM, Dumka LE, Gonzales NA, Knight GK, Michaels ML, et al. Pathways from family economic conditions to adolescents’ distress: Supportive parenting, stressors outside the family, and deviant peers. Journal of Community Psychology. 2002;30:135–152. [Google Scholar]
  8. Bronfenbrenner U. The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press; 1979. [Google Scholar]
  9. Brook JS, Whiteman M, Balka EB, Win PT, Gursen MD. Drug use among Puerto Ricans: Ethnic identity as a protective factor. Hispanic Journal of Behavioral Sciences. 1998;20:241–254. [Google Scholar]
  10. Brooks AJ, Stuewig J, LeCroy CW. A family based model of Hispanic adolescent substance use. Journal of Drug Education. 1998;28:65–86. doi: 10.2190/NQRC-Q208-2MR7-85RX. [DOI] [PubMed] [Google Scholar]
  11. Buriel R, Calzada S, Vasquez R. The relationship of traditional Mexican American culture to adjustment and delinquency among three generations of Mexican American male adolescents. Hispanic Journal of Behavioral Sciences. 1982;4:41–55. [Google Scholar]
  12. Cauce AM, Domenech-Rodriguez M. Latino families: myths and realities. In: Contreras JM, Kerns KA, Neal-Barnett AM, editors. Latino Children and Families in the United States. Westport, Connecticut: Praeger; 2002. pp. 3–25. [Google Scholar]
  13. Camarota SA. Immigration from Mexico: Assessing the impact on the United States. Retrieved July 31, 2004, from the Center for Immigration Studies Web site: http://www.cis.org/articles/2001/mexico/release.html.
  14. Chavez EL, Oetting ER, Swaim RC. Dropout and delinquency: Mexican-American and Caucasian non-Hispanic adolescents. Journal of Clinical Child Psychology. 1994;23:47–55. [Google Scholar]
  15. Cohen J, Cohen P, West SG, Aiken LS. Applied multiple regression/correlation analysis for the behavioral sciences. Mahwah, NJ: Lawrence Erlbaum Associates; 2003. [Google Scholar]
  16. Crick N, Zahn-Waxler C. The development of psychopathology in females and males: Current progress and future challenges. Development and Psychopathology. 2003;15:719–742. [PubMed] [Google Scholar]
  17. Cuellar I, Arnold B, Maldonado R. Acculturation Rating Scale for Mexican Americans-II: A revision of the original ARSMA scale. Hispanic Journal of Behavioral Sciences. 1995;17 (3):275–304. [Google Scholar]
  18. Dishion TJ, Patterson GR, Stoolmiller M, Skinner ML. Family, school, and behavioral antecedents to early adolescent involvement with antisocial peers. Developmental Psychology. 1991;27:172–180. [Google Scholar]
  19. Dumka LE, Lopez VA, Carter SJ. Parenting interventions adapted for Latino families: Progress and Prospects. In: Contreras JM, Kerns KA, Neal-Barnett AM, editors. Latino Children and Families in the United States. Westport, Connecticut: Praeger; 2002. pp. 204–231. [Google Scholar]
  20. Formoso D, Gonzales NA, Barrera M, Jr, Dumka LE. Marital relations and maternal employment:Systemic effects on fathering and child outcome. Journal of Mariage and the Family (in press) [Google Scholar]
  21. Fuligni AJ, editor. New directions for child and adolescent development. San Francisco, CA: Jossey-Bass; 2001. Family obligation and assistance during adolescence: Contextual variations and developmental implications; pp. 61–75. [DOI] [PubMed] [Google Scholar]
  22. Fuligni AJ, Tseng V, Lam M. Attitudes toward family obligations among American adolescents with Asian, Latin American, and European backgrounds. Child Development. 1999;70:1030–1044. [Google Scholar]
  23. Garcia Coll C, Akerman A, Cicchetti D. Cultural influences on developmental processes and outcomes: Implications for the study of development and psychopathology. Development and Psychopathology. 2000;12(3):333–356. doi: 10.1017/s0954579400003059. [DOI] [PubMed] [Google Scholar]
  24. Gil AG, Warheit GJ, Zimmerman RS, Apospori E. Acculturation and delinquent behavior among Cuban-American adolescents: Toward an empirical model. American Journal of Community Psychology. 1993;21:113–125. doi: 10.1007/BF00938210. [DOI] [PubMed] [Google Scholar]
  25. Gonzales NA, Knight GP, Birman D, Sirolli AA. Acculturation and enculturation among Latino youth. In: Maton KI, Schellenbach CJ, Leadbeater BJ, Solarz AL, editors. Investing in children, youth, families, and communities: Strengths-based research and policy. Washington, DC: American Psychological Association; 2004. pp. 285–302. [Google Scholar]
  26. Gonzales NA, Knight GP, Morgan-Lopez AA, Saenz D, Sirolli A. Acculturation and the mental health of Latino adolescents: An integration and critique of the literature. In: Contreras JM, Kerns KA, Neal-Barnett AM, editors. Latino Children and Families in the United States. Westport, Connecticut: Praeger; 2002. pp. 204–231. [Google Scholar]
  27. Gonzales NA, Knight GP, Saenz D. Codebook for the Mexican American Ethnic Pride Scale. Arizona State University Program for Prevention Research; 2005. [Google Scholar]
  28. Heller PL. Familism Scale: a measure of family solidarity. Journal of Marriage and the Family. 1970;32:73–80. [Google Scholar]
  29. Hirschi T. Causes of delinquency. Berkeley, CA: University of California Press; 1969. [Google Scholar]
  30. Kaplan CP, Napoles-Springer A, Stewart SL, Perez-Stable EJ. Smoking acquisition among adolescents and young Latinas: the role of socioenvironmental and personal factors. Addictive Behaviors. 2001;26:531–550. doi: 10.1016/s0306-4603(00)00143-x. [DOI] [PubMed] [Google Scholar]
  31. Keefe SE, Padilla AM, Carlos ML. The Mexican-American Extended Family As An Emotional Support System. Human Organization. 1978;38:144–152. [Google Scholar]
  32. Knight GP, Gonzales NA, Saenz D, Germán M, Deardorff J, Roosa MW, Updegraff KA. The Mexican American cultural values scale for adolescents and adults. 2007 doi: 10.1177/0272431609338178. Manuscript submitted for publication. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mason CA, Cauce AM, Gonzalesa N, Hiraga Y. Adolescent problem behavior: The effect of peers and the moderating role of father absence and the mother/child relationship. American Journal of Community Psychology. 1995;22:723–743. doi: 10.1007/BF02521556. [DOI] [PubMed] [Google Scholar]
  34. Oetting ER, Donnermeyer JF. Primary socialization theory: the etiology of drug use and deviance. Substance Use & Misuse. 1998;33:995–1026. doi: 10.3109/10826089809056252. [DOI] [PubMed] [Google Scholar]
  35. Parke RD, Coltrane S, Borthwick-Duff S, Powers J, Adams M, Fabricius W, Braver S, Saenz D. Assessing father involvement in Mexican-American families. In: Day RD, Lamb ME, editors. Conceptualizing and measuring father involvement. Mahwah, NJ: Lawrence Erlbaum Associates; 2004. pp. 17–38. [Google Scholar]
  36. Phinney JS, Ong A, Madden T. Cultural values and intergenerational value discrepancies in immigrant and non-immigrant families. Child Development. 2000;71:528–539. doi: 10.1111/1467-8624.00162. [DOI] [PubMed] [Google Scholar]
  37. Roosa MW, Morgan-Lopez AA, Cree WK, Specter MM. Ethnic Culture, Poverty, and Context: Sources of Influence on Latino Familias and Children. In: Contreras JM, Kerns KA, Neal-Barnett AM, editors. Latino Children and Families in the United States. Westport, CN: Praeger Publishers; 2002. pp. 27–44. [Google Scholar]
  38. Sabogal F, Marin G, Otero-Sabogal R, Marin BV, Perez-Stable EJ. Hispanic familism and acculturation: What changes and what doesn’t? Hispanic Journal of Behavioral Sciences. 1987;9:397–412. [Google Scholar]
  39. Samaniego RY, Gonzales NA. Multiple mediators of the effects of acculturation status on delinquency for Mexican-American adolescents. American Journal of Community Psychology. 1999;27:189–210. doi: 10.1023/A:1022883601126. [DOI] [PubMed] [Google Scholar]
  40. Seidman E, Allen L, Aber JL, Mitchell C, Feinman J. The impact of school transitions in early adolescence on the self-system and perceived social context of poor urban adolescents. Child Development. 1994;65:507–522. [PubMed] [Google Scholar]
  41. Seidman E, Lambert LE, Allen L. Protective family processes and the transition to junior high school. In: Schellenbach C, Maton K, Leadbetter B, Solarz A, editors. Investing in children, families, and communities: Strengths-based research and policy. Washington, DC: American Psychological Association; 2001. [Google Scholar]
  42. Sirolli A. The relation of Mexican American cultural orientation to familism. Arizona State University; 2004. Unpublished doctoral dissertation. [Google Scholar]
  43. Sommers I, Fagan J, Baskin D. Sociocultural influences on the explanation of delinquency for Puerto Rican adolescents. Hispanic Journal of Behavioral Sciences. 1993;15:36–62. [Google Scholar]
  44. Suárez-Orozco C, Suárez-Orozco MM. Transformations: Immigration, family life, and achievement motivation among Latino adolescents. Cambridge, MA: Stanford University Press; 1995. [Google Scholar]
  45. Toth JF, Jr, Xu X. Ethnic and cultural diversity in fathers’ involvement: A racial/ethnic comparison of African American, Hispanic, and White fathers. Youth & Society. 1999;31:76–99. [Google Scholar]
  46. Tolan PH, Guerra NG, Montaini-Klovdahl LR. Staying out of harm’s way: Coping and the development of inner-city children. In: Wolchik SA, Sandler IN, editors. Handbook of children’s coping: Linking theory and intervention. New York, NY: Plenum Press; 1997. pp. 453–479. [Google Scholar]
  47. U.S. Department of Education. Dropout rates in the United States: 1998. Washington DC: U.S. Government Printing Office; 2000. (NCES 2000–022) [Google Scholar]
  48. Updegraff KA, McHale SM, Whiteman SD, Thayer SM, Delgado MY. Adolescent sibling relationships in Mexican American families: exploring the role of familism. Journal of Family Psychology. 2005;19:512–522. doi: 10.1037/0893-3200.19.4.512. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Valdés G. Con Respecto: Bridging the distance between culturally diverse families and schools. New York, NY: Teachers College Press; 1996. [Google Scholar]
  50. Vega WA, Gil AG, Warheit GJ, Zimmerman RS, Apospori E. Acculturation and delinquent behavior among Cuban American adolescents: toward an empirical model. American Journal of Community Psychology. 1993;21:113–125. doi: 10.1007/BF00938210. [DOI] [PubMed] [Google Scholar]
  51. Vigil JD. Learning from gangs: The Mexican-American experience. ERIC digest. 1997 February; Retrieved July 26, 2004, from http://www.ericfacility.net/databses/ERIC_Digests/ed405157.html.
  52. Vitaro F, Tremblay RE, Kerr M, Pagani L, Bukowski WM. Disruptiveness, friends; characteristics, and delinquency in early adolescent: A test of two competing models of development. Child Development. 1997;68:676–689. [PubMed] [Google Scholar]
  53. West SG, Aiken LS, Krull JL. Experimental personality designs: Analyzing categorical by continuous variable interactions. Journal of Personality. 1996;64:1–48. doi: 10.1111/j.1467-6494.1996.tb00813.x. [DOI] [PubMed] [Google Scholar]

RESOURCES