Cultural Socialization and Ethnic Pride Among Mexican-Origin Adolescents During the Transition to Middle School

Abstract

The relation between cultural socialization and ethnic pride during the transition to middle school was examined for 674 5^th grade students (50% boys; M_age = 10.4 years) of Mexican origin. The theoretical model guiding the study proposes that parent-child relationship quality is a resource in the transmission of cultural values from parent to child and that parental warmth promotes the child's positive response to cultural socialization. Results showed that mother and father cultural socialization predicted youth ethnic pride, and that this relation was stronger when parents were high in warmth. The findings highlight the positive role parent cultural socialization may play in the development of adolescent ethnic pride. Furthermore, findings reveal the role of parent-child relationship quality in this process.

Within ethnic minority families, the maintenance of ethnic pride by individual family members appears to be an important psychological resource in the face of adversity, helping promote healthy academic and psychological adjustment (Berkel et al., 2010; Brook, Whiteman, Balka, Win, & Gursen, 1998; Guilamo-Ramos, 2009; Marsiglia, Kulis, & Hecht, 2001; Smokowski, Buchanan, & Bacallao, 2009; Umaña-Taylor, Diversi & Fine, 2002). During the transition to adolescence, individuals experience potentially stressful cognitive, biological, and peer social changes, as well as contextual changes in the family (e.g., increases in parent-child conflict) and school (e.g. elementary vs middle school academic demands; Eccles et al., 1993). Especially during this challenging transition, a sense of ethnic pride may help minority children successfully cope with these stressful circumstances. Moreover, parents may foster this psychological resource for their children.

With these ideas in mind, the purpose of the present study was to investigate family factors believed to promote the development of ethnic pride during the transition from 5^th to 7^th grade among Mexican-origin youth. Guided by previous theory and research in this area (e.g., Grusec & Goodnow, 1994; Knight, Bernal, Garza, Cota, & Ocampo, 1993; Parke & Buriel, 2006), we evaluated a theoretical model (Figure 1) that predicts father and mother cultural socialization (i.e., teaching children to be proud of their ethnic background and encouraging traditional Mexican or Mexican-American values) will promote the development of adolescent ethnic pride during the transition from childhood to adolescence (path A), after controlling for earlier ethnic pride (path B). The model also predicts that father and mother warmth (i.e., parent expressions of care, encouragement, and helpfulness) will directly foster the development of adolescent ethnic pride (path C). In addition to the hypothesized direct effect of parental warmth on ethnic pride, the model also predicts that parental warmth will moderate the effect of parent cultural socialization on ethnic pride (path D). This hypothesis is based on research showing that positive parenting behaviors provide a context that promotes the child's acceptance of parental values, rules, and regulations (Grusec & Goodnow, 1994; Pratt, Hunsberger, Pancer, & Alisat, 2003). Thus, the model proposes that parental warmth will amplify the positive impact of cultural socialization on the development of ethnic pride. We next discuss each of the pathways in the model in relation to earlier theory and research findings.

Figure 1.

Theoretical model for the development of ethnic pride.

Empirical Findings and Theoretical Approaches

In the following review, we define the key concepts in Figure 1 (i.e., cultural socialization, parental warmth, and ethnic pride) and explain the significance of ethnic pride for the positive development of Mexican origin and other minority youth. Second, we review earlier findings on the topic of cultural socialization, discussing how the present study builds on prior work and overcomes certain limitations of many previous studies. Third, we present an alternative view of the cultural socialization process by integrating broader conceptualizations of parenting with a more culturally specific process model. Finally, we consider some of the limitations in earlier research, which the present study helps to address.

Key Study Concepts

Hughes et al. (2006) identified four types of ethnic-racial socialization: cultural socialization, preparation for bias, promotion of mistrust, and egalitarianism. As noted, the current inquiry focuses specifically on cultural socialization. Umaña-Taylor and Fine (2004) suggest that cultural socialization can include both deliberate (or overt) and implicit (or covert) intents to socialize children concerning positive aspects of their cultural background. In the present study, we assessed overt cultural socialization from parents, which includes direct teaching about being proud of one's Mexican or Mexican-American background and encouraging traditional Mexican or Mexican-American values, such as considering family when making important decisions.

Another key concept in this study is parental warmth, which has been linked to higher levels of ethnic identity (e.g., Caldwell, Zimmerman, Bernat, Sellers, & Notaro, 2002; Gaylord-Harden, Ragsdale, Mandara, Richards, & Petersen, 2007). Parental warmth was assessed using observer ratings of the intensity and frequency of warmth directed by mothers and fathers toward the child during 20-minute discussion tasks. Examples of such behavior include the parent's expressions of care and encouragement toward the child or demonstrations of helpfulness and respect toward the child. This approach to assessing parental warmth has demonstrated adequate reliability and predictive validity in earlier studies of parent-child relationships with both European American and African American families (Conger et al., 2002; Melby & Conger, 2001). The present study extends this methodology to Mexican-origin families.

The outcome of interest in this study is a youth's sense of ethnic pride, which is a particular aspect of ethnic identity that represents positive attitudes toward one's ethnic group (Phinney, 1990). Research shows that feelings of ethnic pride relate to positive development for ethnic minority youth, either by directly promoting competence and well-being or by functioning as a protective factor against risk and adversity. For example, youth with a strong sense of ethnic pride were less likely to intend to smoke cigarettes, as shown in a study by Guilamo-Ramos (2009) with Latino youth (N = 1,538; 7.5% Mexican origin; 6^th-8^th graders) living in the Bronx, NY. In a study of adolescent and young adult Puerto Ricans living in New York (N = 555; M_age = 19), Brook et al. (1998) found that participants' sense of belonging to one's cultural group was negatively associated with drug use. Also, in a study of youth (N = 408; 52% Mexican American; 7^th graders) living in the southwest, Marsiglia et al. (2001) reported that Mexican American participants with greater ethnic pride were less likely to report use of alcohol, tobacco, or marijuana, and also reported receiving fewer offers of drugs.

Ethnic pride also appears to promote more positive self-views and social interactions. For example, in their review of studies with Latino adolescents, Umaña-Taylor et al. (2002) found that ethnic identity was positively linked to self-esteem. Also, Latino immigrant adolescents' (N = 281; 58% Mexican origin; 11-18-year-olds) who were more involved with their culture of origin—an aspect of ethnic pride—experienced higher self-esteem and fewer social problems (e.g., not getting along with peers, getting teased a lot, arguing a lot, getting into fights) and feelings of hopelessness (Smokowski et al., 2009). Both theoretically and empirically, ethnic pride appears to be an important psychological resource that promotes well-being in a variety of domains and across different developmental periods. These findings underscore the importance of increasing understanding of the origins of ethnic pride, such as parental socialization.

Family Cultural Socialization and Ethnic Pride

Consistent with the paths illustrated in Figure 1, earlier studies suggested that parent cultural socialization may promote ethnic pride among Mexican American youth, a link also seen for youth from other minority backgrounds (Berkel et al., 2010; Hughes, Hagelskamp, Way, & Foust, 2009; see Hughes et al., 2006 for review; Knight et al., 2011; Knight, Bernal, et al., 1993; Quintana & Vera, 1999; Rivas-Drake, Hughes, & Way, 2009; Umaña-Taylor & Fine, 2004). Although most cultural socialization research has focused on African American youth (e.g., Hughes et al., 2009; Hughes et al., 2006), here we highlight studies examining the association between cultural socialization and ethnic pride for Mexican-origin youth.

In a study of 45 Mexican American children (age range: 6-10 years), Knight, Bernal, et al. (1993) found that mother's cultural socialization, defined as teaching about Mexican culture, was related to two domains of children's ethnic identity – use of ethnic role behaviors and ethnic preferences. In a more recent study with Mexican American youth from Texas (N = 513; age range: 16-19 years), Umaña-Taylor and Fine (2004) reported that familial cultural socialization, which included both intentional and non-intentional teaching about Mexican American culture, was positively related to ethnic identity achievement, a concept complementing ethnic pride.

Directly paralleling the age range of the current study, Knight et al. (2011) examined the process by which cultural values were passed from parent to youth in a community study of Mexican American families in the southwest (N = 750; 5^th to 7^th grade youth; 74% of mothers & 80% of fathers born in Mexico). They found that mother cultural socialization positively predicted youth ethnic identity (defined by exploration and resolution), but father cultural socialization did not. Knight et al. (2011) suggested that mothers are more likely to be the primary socializing agents of their children's cultural orientation. Father socialization in traditional Mexican American families may involve an indirect process or may depend on the developmental stage of the child so the process becomes more salient in later adolescence. Indeed, some research suggests that females often have a more salient role in passing on cultural traditions González, Umaña-Taylor, & Bámaca, 2006; Phinney, 1990), consistent with Knight et al.'s (2011) findings. Because of the limited evidence on this issue and theoretical expectations, we proposed that cultural socialization by both fathers and mothers will have a positive influence on the development of ethnic pride in early adolescence. In theory, when fathers actively engage in the cultural socialization process, they also should have a positive influence on their children (Parke & Buriel, 2006).

Combining Family Support and Socialization Models

As shown in Figure 1, our theory of the cultural socialization process also takes into account the broader context of parenting. Examining this link during the transition to adolescence is of particular interest, given the developmental trends of increasing parent-child conflict and the continued role parents have on child development (Eccles et al., 1993). Specifically, our theory builds on the suggestion by Hughes et al. (2006) that parent-child relationship quality should be considered as a resource that will enhance the transmission of ethnic values from parent to child. This hypothesis is shown in Figure 1 by path D. This arrow indicates our prediction that parental warmth will amplify or increase the effect of socialization practices on the development of ethnic pride, a statistical interaction effect. The figure also illustrates our expectation that warmth will directly increase ethnic pride, a statistical main effect (path C).

Extensive research has documented that parental warmth, an important family-level process, is linked to positive child development, such as better mental health and academic success for Mexican-origin youth and youth from other ethnic backgrounds (Bámaca-Colbert, Umaña-Taylor, & Gayles, 2012; Conger et al., 2002; DeGarmo & Martinez, 2006; Gonzales et al., 2011; Parke et al., 2004). Research also has shown that warm and supportive parenting is positively associated with children's sense of self-worth (Maccoby, 1992), a concept related to ethnic pride for Latino adolescents (Umaña-Taylor et al., 2002). However, little attention has been given to how such support is related to ethnic identity development and the cultural socialization process. In one study, Knight et al. (2010) found that parental warmth and acceptance was associated with Mexican American cultural values (i.e., familismo, respeto) among Mexican origin youth (60.6% of mothers and 68.6% of fathers born in Mexico). More closely related to ethnic pride, another study showed that general maternal support (i.e., emotional, problem-solving, and moral support) was positively associated with ethnic identity for African American adolescents (Caldwell et al., 2002). Gaylord-Harden et al. (2007) also found that social support (similar to parent warmth) was positively associated with ethnic identity among African American youth. To our knowledge, the link between parent warmth and youth ethnic identity has not been examined for Mexican-origin youth, highlighting the need for replication with this ethnic group. As noted and consistent with the few available findings (Caldwell et al., 2002; Gaylord-Harden et al., 2007), we hypothesized that parental warmth, as measured by observed behaviors toward the child, will be directly and positively related to ethnic pride.

With regard to our hypothesis that parental warmth will amplify the positive influence of cultural socialization on the development of ethnic pride, Grusec and Goodnow (1994) proposed that positive behaviors of parents in general provide a context that leads to child acceptance or internalization of parental values, rules, and regulations. Similarly, we proposed that internalization of values reflecting ethnic pride would be facilitated by parent warmth and a child-centered approach in general. Grusec and Goodnow (1994) suggest that parental warmth is one part of a process expected to promote acceptance of parental priorities. Parental warmth is hypothesized to promote acceptance by increasing the child's desire to be like the parent. The child internally processes the parents' socialization efforts and accepts the validity of these efforts, applying them to their own thinking and behavior. Consistent with this view, Pratt et al. (2003) found that warm parenting moderated the association between parent and adolescent values, so that greater parental warmth was associated with stronger parent-adolescent agreement in terms of their moral values. Based on the work of Grusec and Goodnow (1994) and Pratt et al. (2003), we hypothesized that parental warmth will promote children's positive response to their parents' cultural socialization (Figure 1). Finally, Figure 1 also indicates that we expected ethnic pride during late childhood to predict later ethnic pride during early adolescence (path B). With this variable in the model, we will be conducting a conservative test of the study hypotheses by predicting relative change in ethnic pride from childhood to early adolescence.

Limitations in Earlier Studies

In addition to theoretical and empirical contributions to the study of the development of ethnic pride, the current report also extends earlier research by addressing common methodological limitations in many previous studies. First, many earlier studies are limited to reports by a single informant. For example, studies that have examined cultural socialization and ethnic pride typically have used youth-reports for both measures (see Hughes et al., 2006; Rivas-Drake et al., 2009; Umaña-Taylor & Fine, 2004), although a few investigations have avoided this limitation (e.g., Hughes et al., 2009; Knight et al., 2011; Knight, Bernal, et al., 1993; Quintana & Vera, 1999). The problem with using only youth reports of their parents' cultural socialization practices is that certain practices may be more salient depending on the child or adolescent's own level of ethnic identity or ethnic pride, thus creating a bias in the report and blurring the directionality of influence. In addition, almost no studies have distinguished father and mother cultural socialization practices (see Knight et al., 2011 for an exception). To address these limitations, the present study uses both father and mother reports of their cultural socialization practices toward their child and child reports of ethnic pride. Both father and mother effects are considered within a single set of analyses, allowing us to examine their independent effects on ethnic pride. Finally, most studies have used cross-sectional research designs (e.g., Hughes et al., 2009; Quintana & Vera, 1999; Rivas-Drake et al., 2009; Umaña-Taylor & Fine, 2004), which cannot evaluate whether cultural socialization at an earlier point in time is associated with the development of ethnic pride across time (see Knight et al., 2011 for an exception). The present prospective study allows consideration of the temporal ordering of hypothesized causes and effects.

The Present Study

The present study evaluated three hypotheses regarding Mexican-origin youth making the transition from 5^th to 7^th grade:

• Hypothesis 1

  Father and mother cultural socialization will positively and additively influence children's later ethnic pride and change in ethnic pride during the transition from late childhood to early adolescence.

• Hypothesis 2

  Greater father and mother warmth, as measured by observer ratings, will positively influence later ethnic pride and change in ethnic pride during the transition from late childhood to early adolescence.

• Hypothesis 3

  Higher levels of parental warmth, as measured by observer ratings, will strengthen the positive association between parent cultural socialization and change in ethnic pride during the transition from late childhood to early adolescence.

Analyses also included child gender, family income, and parent education as control variables. One study of adolescent youth found that girls reported more cultural socialization from their parents (Huynh & Fuligni, 2008). Thus, we controlled for child gender. Also, socioeconomic status (SES) is related to the socialization practices of parents, so we controlled for family income and parent education (Conger, Conger, & Martin, 2010).

Method

Participants and Procedures

Participants in this report included Mexican-origin families with a child (N = 674, M_age = 10.4 years, 50% female) who attended the fifth grade in public (95%) or Catholic schools (5%). Children of Mexican origin were drawn at random from student rosters provided by the school districts in one of two cities in a metropolitan area of Northern California, one with 466,488 total population (22.6% of Mexican origin) and one with 55,468 total population (43.9% of Mexican origin; 2010 Census Summary File, 2010). Families of these children were then recruited by telephone or, for cases without a listed phone number, by a recruiter who went to their home. Of eligible families, 68.6% agreed to participate in the study. All family members were of Mexican origin as determined by their ancestry and self-identification as being of Mexican heritage.

In the final participating group, 88% of fathers, 84% of mothers, and 28% of children were born in Mexico. Either two-parent (82%) or single-parent (18%) families were eligible to participate, but all children had to be living with their biological mother. In two-parent families, the father had to be the child's biological father at study initiation. In current analyses, we included data collected during fifth grade (late childhood) and seventh grade (early adolescence). Eighty-six percent of the original families participated during seventh grade. Attrition analyses showed that families who withdrew from the study were not significantly different from those who remained in the study on demographic characteristics (i.e., income, parent education, gender) or on study variables (p < .05), after adjusting for multiple comparisons.

Trained research staff interviewed the participants in their homes using laptop computers. All interviewers were bilingual, and most were of Mexican heritage. Interviews were conducted in Spanish or English based on the preference of each participant. Incentives to participate included $200 for two-parent families and $125 for one-parent families.

Parents and target children also participated in videotaped interaction tasks in their homes. Father and child and mother and child each completed a separate 20-min discussion task, randomly counterbalanced in terms of order. The interviewers provided instructions, set up and started the video equipment, gave participants a set of cards containing discussion questions, and then left the room so they could not hear the video-recorded discussion. The questions asked the caregiver and child to discuss a range of issues in their daily lives – from pleasurable activities they do together to how they handle conflicts and disagreements. Videotaped discussions occurred in a location that provided as much privacy as possible.

Prior to rating the parent-child interactions, project observers received several weeks of training on rating family interactions. Before observing video interactions, coders had to independently rate pre-coded interaction tasks and achieve at least 90% agreement with the “gold standard” ratings of those tasks. To evaluate interobserver reliability, 20% of the tasks were randomly selected and rated by a second observer.

Measures

To avoid any biases produced by having the same reporter for two adjacent variables in the theoretical model (Figure 1), each measure was generated by a different source of information – parents, children, and trained observers.

Cultural socialization

When children were in fifth grade, parents were asked how often they tried to influence their children's cultural and ethnic understanding by teaching them to be proud of their Mexican heritage. Mothers and fathers responded to nine questions assessing their own cultural socialization practices toward the target child. Five items were adapted from a measure developed by Bernal and Knight (e.g., How often do you tell your child about successful Mexican Americans who live in your community? How often do you encourage your child to speak Spanish?; Bernal & Knight, 1993;Knight, Bernal, et al., 1993; Knight, Cota, & Bernal, 1993). Four additional items were adapted from a scale created by Knight et al. (2011) in a later study (e.g., “How often do you tell your child that his or her behavior reflects on the family?”). All items were selected to be conceptually consistent with the operational definition of overt cultural socialization. Responses were made on four-point Likert-type scales, ranging from 1(almost never) to 4 (almost always), with higher scores indicating more frequent socialization practices (mother α = 0.77, father α = 0.81; see Table 1).

Table 1. Intercorrelations and Descriptive Statistics for Latent Variables.

	Grade	1	2	3	4	5	6
Mother
1. Cultural Socialization	5th	---
2. Warmth	5th	.15**	---
Father
3. Cultural Socialization	5th	.21***	.08	---
4. Warmth	5th	.05	.22***	.11+	---
Adolescent
5. Ethnic Pride	5th	.17***	.18**	.05	.01	---
6. Ethnic Pride	7th	.16**	.11*	.07	-.06	.34***	---
M		3.01	4.60	2.94	4.12	3.58	3.64
SD		.51	1.30	.54	1.26	.44	.46
α		.77	.71	.81	.73	.75	.86

N = 674.

⁺p < .10

^*p < .05

^**p < .01

^***p < .001.

These scale items were converted into three parcels by randomly assigning items to create three indicators for the latent factor of cultural socialization. Previous research supports parceling to decrease measurement error and increase reliability of the latent variable (Little, Cunningham, Shahar, & Widaman, 2002). Table 1 provides the correlations among latent variables along with coefficient alphas, means and standard deviations for the measures used in the study. Standardized factor loadings for mother socialization ranged from .69 to .73 and from .72 to .76 for father socialization.

Observed mother and father warmth toward child

Parent warmth was assessed using observer ratings of parents' behaviors to the child in the videotaped interaction task during fifth grade. Parent behavior was assessed using the Iowa Family Interaction Rating Scales (IFIRS; Melby & Conger, 2001), which were adapted for use with Mexican origin families in this study. The IFIRS have demonstrated reliability and validity with European American families living in Iowa and African American families living in Georgia and Iowa (Conger, Ge, Elder, Lorenz, & Simons, 1994; Conger et al., 2002). For the present study, the IFIRS were adapted to reflect Mexican origin family practices, based on focus groups made up of individuals from the Mexican American communities in the area where the study was conducted. Observing and rating family interactions has several scientific and practical benefits. First, observational measures overcome many of the methodological biases produced by a singular reliance on self-reports by family members (Caspi, 1998; Kagan, 1998). Second, observer ratings also generate family process measures that have been particularly effective in predicting and documenting developmental change over time (Conger, Lorenz, & Wickrama, 2004; Scaramella, Conger, & Simons, 1999). Third, observational measures of family interactions provide information that readily translates into intervention targets in the design of more effective prevention programs; as such, they provide particularly sensitive indicators of family change in the evaluation of program effectiveness (Rueter, Conger, & Ramisetty-Mikler, 1999; Spoth, Redmond, & Shin, 2000).

Observers rated the intensity and frequency of warmth directed by the caregiver to the child using a composite of two separate rating scales: warmth and support, and prosocial behaviors. Warmth and support refers to the expression of care, concern, support, or encouragement toward the child. Prosocial behaviors are demonstrations of helpfulness, sensitivity, cooperation, sympathy, and respect toward the child (see Melby & Conger, 2001 for more detailed descriptions). For both dimensions, observers rated parents on a scale that ranged from 1 (behavior is not at all characteristic of the parent) to 9 (behavior is mainly characteristic of the parent). Interobserver reliability was examined with intraclass correlations and was .65 for mother warmth and .65 for father warmth. Standardized indicator loadings for these measures ranged from .61 to .92. The alpha coefficients were .71 (mothers) and .73 (fathers; see Table 1).

Ethnic pride

Mexican American ethnic pride was assessed using six items, reported by the adolescent in 5^th and 7^th grades. Three items come from a scale that has been used successfully in other studies with Mexican American adolescents (Berkel et al., 2010; Knight et al., 2010): “You have a lot of pride in your Mexican roots,” “You like people to know that your family is Mexican or Mexican-American,” and “You feel proud to see Latino actors, musicians and artists being successful.” The other three items were adapted from the Ethnic Affirmation and Belonging subscale from the Multigroup Ethnic Identity Measure developed by Phinney (1990): “You feel good about your cultural or ethnic background,” “You are happy that you are Mexican American,” and “You feel a strong attachment towards your own ethnic group.” The items were rated on a four-point Likert-type scale, ranging from 1 (not at all true) to 4 (very true), with higher scores indicating higher levels of ethnic pride (α = 0.75 in 5^th grade; α = 0.86 in 7^th grade). These scale items were converted into three parcels by randomly assigning items to create indicators for the latent factor of ethnic pride. The standardized loadings for the indicators of ethnic pride ranged from .66 to .86.

Control variables

Gender (0 = girl, 1 = boy), mother-reported family income in $5,000 increments (M = $30,000-$35,000, SD = $15,000-$20,000, range: less than $5,000-more than $95,000), and parent education (M = 9.3 years, SD = 3.3 years) were used as control variables. Preliminary analyses examined whether there were any significant interactions between the control variables (education, income, gender) and any independent variables (cultural socialization, warmth) in predicting the dependent variable (ethnic pride). None of these interactions was statistically significant.

Results

Preliminary Analyses and Correlations

The theoretical model was tested within a structural equation modeling (SEM) framework using Mplus Version 6 (Muthén & Muthén, 1998-2010). SEM allows the simultaneous testing of multiple hypothesized relations among latent factors. To deal with missing data, we used full information maximum likelihood estimation to fit models directly to the raw data. This procedure produces less biased and more reliable results compared with conventional methods of dealing with missing data, such as listwise or pairwise deletion (Allison, 2003; Widaman, 2006). Model fit was considered adequate if the likelihood ratio (χ²/df) was less than 2.0, the comparative fit index (CFI) was above .95 (Bentler, 1990), the Tucker-Lewis index (TLI) was above .95 (Tucker & Lewis, 1973), and the root mean square error of approximation (RMSEA) was below .05 (Browne & Cudeck, 1993). We first identified a measurement model that examined the degree to which each manifest variable loaded onto the appropriate latent factor.

We tested for measurement invariance across mothers and fathers on the cultural socialization and warmth latent variables and across time for ethnic pride. We found support for strict measurement invariance, which entails invariant factor loadings, measurement intercepts, and unique factor variances (except that two ethnic pride parcel residual variances were allowed to vary). This measurement model demonstrated acceptable fit: CFI = .99, TLI = .99, and RMSEA = .02.

Table 1 provides the zero-order correlations among the latent variables. Consistent with the conceptual model, mother cultural socialization (r = .16) and warmth (r = .11) in fifth grade were significantly associated with the child's ethnic pride in 7^th grade. For fathers, however, neither cultural socialization nor warmth was significantly associated with ethnic pride. These results underscore the importance of considering mothers and fathers separately in subsequent analyses.

Testing the Theoretical Model

We first examined main effects in a latent path regression model with good fit: CFI = .99, TLI = .99, and RMSEA = .02 (see Model 1 in Table 2). Ethnic pride (in 7^th grade) was regressed on father cultural socialization, mother cultural socialization, father warmth, mother warmth, and ethnic pride in fifth grade. We tested for differences in effects between mothers and fathers by constraining the mother and father cultural socialization paths to be equal, as well as mother and father warmth paths to be equal. The resulting model showed a non-significant change in model fit, Δ χ² (2) = 5.37, p > .05, thus the more parsimonious constrained path model was used. Results replicated findings by Knight et al. (2011) when predicting ethnic identity; mother cultural socialization predicted later ethnic pride, B = 0.07, SE = 0.03, p < .01 (see Model 1 in Table 2). Father cultural socialization also predicted ethnic pride at the same magnitude. Contrary to Hypothesis 2, mother and father warmth did not predict ethnic pride. As a next step, we controlled for child gender (0 = girl, 1 = boy), family income, and parent education (see Model 2 in Table 2), and all significant paths remained significant (and of similar magnitude).

Table 2. Comparison of Latent Regression Models Predicting 7^th Grade Ethnic Pride.

	1	2
Model fit
χ2	135.795*	228.402 ***
df	106	143
CFI	0.988	0.968
TLI	0.987	0.962
RMSEA	0.021	0.045
	B (S.E.)	B (S.E.)
Latent Regressions
Mother Cultural Socialization	0.07 (0.03) a *	0.08 (0.03) c *
Mother Warmth	−0.02 (0.02) b	−0.03 (0.02) d
Father Cultural Socialization	0.07 (0.03) a *	0.08 (0.03) c *
Father Warmth	−0.02 (0.02) b	−0.03 (0.02) d
Ethnic Pride 5th	0.38 (0.06) ***	0.40 (0.06) ***
Family Income	---	0.01 (0.01) *
Parent Education	---	−0.01 (0.01)
Child Gender 1	---	−0.07 (0.03) *

Note. Non-standardized coefficients shown. CFI = comparative fit index; TLI = Tucker–Lewis Index; RMSEA = root mean squared error of approximation.

¹0 = girl; 1 = boy.

^a,b,c,dPaths constrained to be equal.

^*p < .05

^**p < .01

^***p < .001.

Our first question for this “main effects only” approach was whether cultural socialization was associated with ethnic pride. Based on the final path model (see Model 2 in Table 2), mother and father cultural socialization had significant effects on rank order change in ethnic pride from 5^th to 7^th grades, B = 0.08, SE = 0.03, p < .05. Our second question was whether observed parent warmth was associated with ethnic pride. As shown in Table 2, neither mother nor father warmth had a significant effect on ethnic pride, B = -0.03, SE = 0.02, p > .05.

Our final question was whether the relation between parent cultural socialization and ethnic pride was moderated by parent warmth. We tested the hypothesized interactions using a latent moderated structural equation (LMS) technique (Klein & Moosbrugger, 2000; Little, Bovaird, & Widaman, 2006), following standard procedures recommended by Muthén and Muthén (1998-2010). To run the interaction term in this default LMS program, there is no need for residual centering and orthogonalizing items (Weiss, 2010). Typical fit indices are not available in Mplus for LMS models, but we compared the Akaike information criterion (AIC; Akaike, 1973) values, with lower values demonstrating improvement in model fit. As recommended by Klein and Moosbrugger (2000), we also compared the log likelihood values and number of parameters from a null interaction model to one that estimated the interaction effect. We added the cultural socialization x warmth interaction latent factors for both mothers and fathers to the regression equation, with mother and father interaction effects constrained to be equal. Constraining the father and mother interaction paths to be equal allowed us to test whether the magnitude of the interaction effects was equivalent, as demonstrated by significant improvement in fit or non-significant worsening of fit. The addition of the constrained interaction paths to the LMS model showed an improvement in fit according to the AIC (Δ AIC = -3.176; compared with the Null LMS model in Table 3). A log likelihood test also demonstrated that including the interaction latent factors significantly improved model fit, Δ χ² (1) = 5.176, p < .05. Relaxing the equality constraint on the interaction terms resulted in a non-significant improvement in fit. Thus, the constrained interaction latent effects were retained, given superior fit and parsimony.

Table 3. LMS Model predicting 7^th grade ethnic pride.

	Null LMS	LMS
Model fit
AIC	23060.544	23057.368
− 2 log likelihood	22932.544	22927.368
# of parameters	64	65
	B (S.E.)	B (S.E.)
Latent regressions
Mother Cultural Socialization	0.08 (0.03) a *	0.09 (0.03) c **
Mother Warmth	−0.03 (0.02) b	−0.04 (0.02) d +
Mother Cultural Socialization × Mother Warmth	0.00 (0.00)	0.11 (0.05) e *
Father Cultural Socialization	0.08 (0.03) a *	0.09 (0.03) c **
Father Warmth	−0.03 (0.02) b	−0.04 (0.02) d +
Father Cultural Socialization × Father Warmth	0.00 (0.00)	0.11 (0.05) e *
Ethnic Pride 5th	0.39 (0.06) ***	0.39 (0.06) ***
Family Income	0.01 (0.01) *	0.01 (0.01) **
Parent Education	−0.01 (0.01)	−0.01 (0.01)
Child Gender 1	−0.07 (0.03) *	−0.09 (0.03) **

Note. Non-standardized coefficients shown.

¹0 = girl; 1 = boy.

^a,b,c,d,ePaths constrained to be equal. AIC = Akaike's Information Criterion

⁺p < .10

^*p < .05

^**p < .01

^***p < .001.

The final LMS model includes mother and father warmth main effects, mother and father cultural socialization main effects, a father cultural socialization x father warmth interaction effect, a mother cultural socialization x mother warmth interaction effect, prior (i.e., 5^th grade) ethnic pride, and control variables (gender, family income, parent education).

As shown in the second column of coefficients in Table 3, we found a significant interaction between father cultural socialization and father warmth in predicting ethnic pride, as well as a significant interaction between mother cultural socialization and mother warmth in predicting ethnic pride, B = 0.11, SE = 0.05, p < .05 (see LMS Model in Table 3). Based on a calculation technique described by Preacher, Curran, and Bauer (2006), examination of the simple slopes analysis revealed that at higher levels of father warmth (i.e., one SD above the mean), father cultural socialization was significantly and positively associated with later ethnic pride, B = 0.19, SE = 0.06, p < 0.01; whereas, at lower levels of warmth (i.e., one SD below the mean), the association between father cultural socialization and later ethnic pride was not significant, B = 0.00, SE = 0.05 (see Figure 2a). At the mean of father warmth, cultural socialization also was positively associated with later ethnic pride, B = 0.09, SE = 0.03, p < .01. Similar results were found for mothers (see Figure 2b). Figures 2a and 2b also show that children who experienced low levels of both warmth and cultural socialization demonstrated relatively high levels of ethnic pride. We consider this unexpected finding in the following discussion.

Figure 2.

The interaction between parent cultural socialization and parent warmth in predicting ethnic pride for Mexican origin adolescents, displayed separately for fathers and mothers. a. Fathers reported on their cultural socialization practices (5^th grade), observers rated father warmth (5^th grade), and adolescents reported on their own ethnic pride (7^th grade). b. Mothers reported on their cultural socialization practices (5^th grade), observers rated mother warmth (5^th grade), and adolescents reported on their own ethnic pride (7^th grade).

Discussion

The present study tested a theoretical model proposing that parental warmth and support is a resource in the transmission of cultural values from parent to child. Guided by earlier theory and research on family socialization (e.g., Grusec & Goodnow, 1994; Knight, Bernal, et al., 1993; Parke & Buriel, 2006), this study extended the contemporary conceptualization of parenting to a cultural socialization framework, building on earlier research related to the direct association between family cultural socialization and the development of child ethnic identity (Hughes et al., 2006; Knight et al., 2011; Knight, Bernal, et al., 1993; Umaña-Taylor & Fine, 2004). We proposed that cultural socialization is an important process for children and adolescents in ethnic minority families (Hughes et al., 2006), including families of Mexican heritage (Umaña-Taylor & Fine, 2004), because it contributes to the development of ethnic identity which appears to promote well-being and other aspects of positive youth development.

In the present study, we used multi-informant reports from fathers, mothers, children, and trained observers to reduce biases associated with shared method variance and provide a strong test of study hypotheses. We also included both father and mother warmth and cultural socialization in our models. Finally, we used a longitudinal research design to control for prior levels of the primary outcome variable and, therefore, predict rank-order change in child ethnic pride from 5^th to 7^th grade. The study hypotheses were generally supported, but some unexpected results highlight the need for future research on the processes through which ethnic pride develops in the context of Mexican American parent-child relationships.

Our first hypothesis was that father and mother cultural socialization would positively predict children's later ethnic pride. Although other studies have found evidence for this association (e.g., Hughes et al., 2006; Knight, Bernal, et al., 1993; Umaña-Taylor & Fine, 2004), replicating the basic finding is important given its centrality to theoretical reasoning about the socialization of ethnic identity in minority group populations. Our results showed that both mother and father cultural socialization predicted development of ethnic pride in 7^th grade. This study provided a relatively strict test of the main effects of parental cultural socialization given that we controlled for child ethnic pride at an earlier time point and used multiple reporters for the measures of interest. The present findings are consistent with the hypothesis that mother and father cultural socialization practices independently influence the development of child ethnic pride, suggesting that each parent serves as a distinct socializing agent in Mexican origin families.

Knight et al. (2011) found that only mothers' cultural socialization was significantly and positively related to child ethnic identity (exploration and resolution), whereas we found that both mothers' and fathers' cultural socialization were positively and significantly related to ethnic pride (affirming, or feeling positively about, one's identity as a Mexican American). The different outcomes examined by Knight et al. (2011) and in the current study may be the source of different effects for mothers and fathers. Ethnic exploration and resolution are distinct from affirmation, which is an “evaluation largely informed by how one believes the group is perceived by others” (Umañ a-Taylor, Vargas-Chanes, Garcia&Gonzales-Backen, 2008, p.120). Thus, mothers and fathers may equally impact ethnic affirmation but not exploration and resolution.

Knight et al. (2011) may have found that only mother ethnic socialization predicted ethnic identity (exploration and resolution) because mothers may have a more direct impact than do fathers on their children's active pursuit in learning about their ethnic identity and having a strong sense of what their ethnicity means to them. For example, mothers typically have a more salient role in passing on cultural traditions (González et al., 2006; Phinney, 1990). The process of engaging in cultural traditions may be more closely aligned with learning about one's culture (exploration) and having a deeper understanding of what one's ethnicity means to them (resolution). Perhaps the stronger role of mothers in passing on cultural traditions accounts for why Knight et al. (2011) found that cultural socialziation from mothers, but not fathers, predicted children's ethnic exploration and resolution identity. The findings in the present study suggest that instilling a sense of pride in one's ethnicity likely comes from both mothers and fathers. These findings corroborate growing evidence of the importance of parenting processes by fathers for child development (Parke & Buriel, 2006). Future research should continue to explore the father's role in the cultural socialization process of youth at different developmental stages and for different types of ethnic understanding by youth.

Our second hypothesis was that greater father and mother warmth, as measured by observer ratings of behavior in parent-child interactions, would positively predict ethnic pride. Although the zero-order correlations indicated that mother warmth predicted later ethnic pride (r = .11, p < .05), this relation was no longer significant when we included other covariates, such as prior levels of ethnic pride. Neither father nor mother warmth predicted later ethnic pride. Little prior research has examined this relation and, to our knowledge, only one study has examined father warmth and support (see Umaña-Taylor & Guimond, 2010 for an exception). We did find two studies that reported a positive link between youth-reported family support and youth-reported ethnic identity in two different groups of African American adolescents (Caldwell et al., 2002; Gaylord-Harden et al., 2007). However, the participants in the Caldwell et al. (2002) study were in twelfth grade, perhaps altering parental support dynamics compared to the fifth grade participants in the present study. Also, our measure of parenting was an observer-based measure of behaviors in parent-child interactions, whereas the Caldwell et al. (2002) and Gaylord-Harden et al. (2007) studies used youth-reports of perceived mother and family support, respectively, raising the possibility that their findings were biased by shared method variance. Umaña-Taylor and Guimond (2010) found that adolescent-reported father support positively predicted ethnic identity exploration (but not resolution or affirmation) among Latino adolescent boys. Future research should further investigate whether youth who perceive their parents as warm and supportive are more likely to have higher levels of ethnic pride and explore how this relation varies from childhood to adolescence and by family and socio-cultural context.

Our third hypothesis was that increased parental support would strengthen the positive association between cultural socialization and child ethnic pride. In the work by Grusec and Goodnow (1994) on parent discipline socialization, and the work by Pratt et al. (2003) on moral value socialization, researchers proposed that parent warmth promotes acceptance, increasing the child's desire to comply with the parent's socialization efforts. We find support for this socialization process with regard to parental cultural socialization. Without considering the interaction with warmth, cultural socialization predicted ethnic pride (B = .08). In the context of high warmth, however, cultural socialization more strongly predicted ethnic pride (B = .19). The results show that in the context of high warmth, parent cultural socialization is more strongly associated with child ethnic pride than is the case when the broader measure of parenting context is not taken into account. The association is moderate in the context of average warmth. Finally, there is no association between cultural socialization and ethnic pride in the context of low warmth.

We note that youth with parents who exhibited low warmth had healthy levels of ethnic pride regardless of the degree of cultural socialization and demonstrated even higher ethnic pride than children who experienced high warmth but little cultural socialization. This finding was not predicted and seems somewhat counterintuitive. We expect that it may be accounted for in at least two ways. First, if highly warm parents do not place an emphasis on cultural issues, as would be the case for parents who engage in little cultural socialization, children may emulate this lack of interest and pay little attention to dimensions of ethnic pride. Thus, their tendency to share similar beliefs and values with their highly warm parents would actually lead to lower levels of concern about ethnicity as reflected in the measure of ethnic pride. Second, given the salience of peer networks in adolescent development and ethnic identity (Kiang & Fuligni, 2009; Rumbaut, 2005), youth experiencing low warmth from parents may retreat to their peers for support and ethnic affirmation. Thus, their ethnic pride may be less affected by the warmth or cultural socialization of their parents and instead be fostered by peers. Future work should attempt to replicate the results of the current study and also extend this research by exploring interactions between peer and parent support and their impact on ethnic pride among Mexican-origin youth.

Caution should be taken however, when interpreting the interactions between cultural socialization and parental warmth. Our measure of socialization assesses overt cultural socialization as reported by fathers and mothers, and thus does not reflect the influence of covert parental socialization practices or socialization by peers and extended family members. Studies suggest that ethnic identity is defined by context, varying by the ethnic composition of the neighborhood (Phinney, 1990). Furthermore, cultural socialization practices are embedded in ecological contexts, such as the availability of community resources in promoting cultural practices. In addition, for Mexican immigrant mothers in particular, extended family can be an important source of cultural socialization (Umaña-Taylor & Bámaca, 2004).

The present study has several limitations. For example, we did not examine whether generational status or acculturative processes in families impacted their cultural socialization practices, as some studies have shown (e.g., Knight et al., 2011; Umaña-Taylor & Fine, 2004). However, research has also emphasized that, for youth from immigrant backgrounds, family has a consequential role in fostering ethnic identity regardless of different acculturation patterns between the parents and youth (Rumbaut, 2005). Nonetheless, future studies should consider acculturative and generational differences in the cultural socialization practices of families. For example, Knight et al. (2011) found that mothers born in Mexico showed higher levels of ethnic socialization. Acculturative processes, approximated by country of birth in this case, may impact parents' childrearing practices, including the extent to which parents engage in cultural socialization.

Future research should also go beyond the specific dimensions of parenting examined in this study. Examining other parenting dimensions, such as control, may yield a more nuanced perspective on how parenting may differentially relate to development among Mexican-origin and Latino families (Bámaca-Colbert, Gayles, & Lara, 2011; Domenech-Rodríguez, Donovick, & Crowley, 2009). Also, future research should test whether the present findings hold when child and parent reports of parenting are used instead of observer ratings.

Despite these limitations, we believe that the results have important implications for future research on the ethnic identity development of Mexican origin adolescents and, more specifically, on the role parents play in this process. Previous research emphasizes the importance of ethnic pride in promoting academic success, mental health, and general well-being from childhood to adulthood – suggesting that ethnic pride is an important source of resilience for children often exposed to very stressful or risky environments (Berkel et al., 2010; Brook et al., 1998; Guilamo-Ramos, 2009; Marsiglia et al., 2001; Smokowski et al., 2009; Umaña-Taylor et al., 2002). Of increasing interest in the research literature and in practice is how parents can foster resilience-promoting dispositions, beliefs, and behaviors in children. Earlier findings and the results of this study suggest that parent cultural socialization may be a key variable for how ethnic pride develops in children. We can also examine how the cultural socialization process relates to other domains of parenting, and provide further insights into how parent socialization functions in domains such as ethnic pride. Although the processes studied here were investigated among families of Mexican-descent, we expect the findings will extend to other ethnic minority families, a proposition that needs to be tested in research with other ethnic minority children and families. The findings from this study support the need for further research on parent-child relationship quality and its effects on the influence of cultural socialization on ethnic pride. The findings also indicate the possible promise of future studies on other parent-child relationship factors that may mediate or moderate the impact of cultural socialization on ethnic pride.