Mexican American 7^th Graders’ Future Work and Family Plans: Associations with Cultural Experiences and Adjustment

Abstract

We describe Mexican American 7^th graders’ expectations for future work and family roles and investigate links between patterns of future expectations and adolescents’ cultural experiences and adjustment. Adolescents participated in home interviews and a series of seven nightly phone calls. Five unique patterns of adolescents’ future expectations were identified (N = 246): Career Oriented, Independent, Family Oriented, Early, and Inconsistent. Career Oriented adolescents had the highest socioeconomic status and contact with the U.S. (e.g., generation status) whereas Family Oriented adolescents had the lowest. Cultural orientations, values, and involvement also varied across groups. For example, Career Oriented adolescents reported significantly higher familism values compared to Inconsistent adolescents. Clusters also differed on adjustment: Career Oriented and Family Oriented adolescents reported higher parental warmth and less risky behavior compared to Independent and Inconsistent adolescents. Findings underscore the multi-faceted nature of adolescents’ future expectations and the diversity in cultural experiences among Mexican origin youth.

A young and rapidly increasing population, large cohorts of Mexican American adolescents who are setting goals now are going to be transitioning to adulthood in the next several decades (Pew Hispanic Center, 2006). Adolescents’ expectations for work and family roles are a critical part of understanding pathways to role attainment in adulthood (Markus & Nurius, 1986; Miller & Brickman, 2004). Though research has documented correlates of career expectations and of family expectations separately (e.g., East, 1998), our interest in this study was to understand how patterns of adult role expectations are associated with individual and contextual characteristics. We used cross-sectional data from young Mexican American adolescents to describe unique patterns of future expectations, and to explore these patterns further by linking them to adolescents’ cultural experiences and adjustment. Studying cultural and adjustment correlates of Mexican American young adolescents’ future expectations provides rich descriptive information about this population and their projected pathways.

The specific goals of this study were threefold. The first goal was to investigate diverse patterns of Mexican American adolescents’ future plans based upon expectations for age to move out of parents’ home, age to get married, educational attainment, and career aspiration. The second and third goals were to examine associations between patterns of adult role expectations and adolescents’ cultural experiences (i.e., orientations, values, and involvement) and adjustment (i.e., depressive symptoms, risky behavior, and parent-adolescent warmth), respectively. The overarching purpose of this study was descriptive, such that our aim was to describe the unique patterns that exist among adolescents’ expectations for adult roles as well as the correlates of those unique patterns.. Further, the cross-sectional nature of our sample led us to consider future expectations as both outcomes and predictors, and we vary our language to reflect this.

Patterns of Future Expectations among Mexican American Adolescents

Experiences in childhood and adolescence can influence future plans and goals that, in turn, can influence pathways to adult role attainment (Crosnoe & Elder, 2004). However, experiences, future plans, and pathways may differ across and within cultural groups (Greenfield, Keller, Fuligni, & Maynard, 2003). For example, “emerging adulthood” in the U.S. is characterized by role exploration, risky activities, and extended education (Arnett, 2000; Cohen, Kasen, Chen, Hartmark, & Gordon, 2003). However, role exploration and extended education are more common in modern industrialized cultures and for individuals in advantaged socioeconomic positions (Shanahan, 2000). Mexican American adolescents, particularly those who have less access to resources via their socioeconomic status and contact with the U.S., may hold beliefs and expectations about transitioning to adult roles that differ from U.S. adolescents of other racial/ethnic backgrounds (Arnett & Eisenberg, 2007). We investigated these ideas with a sample of Mexican American young adolescents. This age period is significant because cognitive skills necessary to think about and plan for future events are maturing (Keating & Sasse, 1996) and control over important decisions about school, relationships, and future careers is increasing. Research indicates that future planning can play a long-term protective role and that its importance is heightened during transition periods, such as early adolescence (McCabe & Barnett, 2000).

An important contribution of this study is our focus on patterns of adult role expectations. Pattern-analytic strategies (Magnusson, 1998) are necessary to capture the multiple ways that adolescents organize their plans for education, career, and family. Knowing the correlates of high versus low educational expectations, for example, becomes more meaningful when it is in the context of also knowing about an adolescents’ family and career goals. To aid our interpretation of emergent patterns, we considered two dimensions along which adolescents’ future plans might vary: family- versus individualistic-orientation, and congruence versus incongruence. For example, a family-orientation could include plans to move out of parents’ home just before entering marriage while an individualistic-orientation could include plans to move out of parents’ home at an early age and to delay marriage for several years. Congruent plans would be logical and realistic, such as aspiring to become a doctor and planning to obtain a professional degree, whereas incongruent plans would be less so (e.g., plans to marry before completing high school).

Links between Patterns of Future Expectations and Cultural Experiences

Cultural experiences may play an important role in Mexican American adolescents’ transitions to adulthood because they must navigate both the majority culture of the U.S. and their culture of origin. Many researchers have noted that culture is particularly important for understanding ethnic minority adolescents’ future expectations because the majority culture and culture of origin may differ in their emphasis on autonomy versus relatedness in adult roles (e.g., Greenfield et al., 2003). For example, spending several years in young adulthood living independent of parents or a spouse may be praised in a culture that values autonomy, but criticized in a culture which stresses the importance of proximal family relationships. To understand this complex issue, we examined several indicators of cultural experiences and we focused on understanding within-group differences in these indicators. In other words, we examined whether adolescents’ cultural orientations, values, and involvement differed across diverse patterns of future plans.

Cultural orientation

One way to measure cultural experiences is to assess orientations to U.S. mainstream and Mexican culture. Some work suggests that mainstream-oriented adolescents, when measured by network preferences, have higher educational goals than adolescents who are Mexican-oriented (e.g., Reyes, Kobus, & Gillock, 1999). Other research, however, finds that time spent living in the U.S. is negatively associated with adolescents’ educational goals (e.g., St-Hilaire, 2002). Socioeconomic status (SES) confounds may explain these discrepant findings. When measured by language and network preferences, mainstream orientation is associated with higher SES (Cuéllar, Arnold, & Maldonado, 1995). Yet, if upward socioeconomic mobility has not occurred as a result of years spent in the U.S., adolescents’ educational goals may diminish (Negy & Woods, 1992). These issues underscore the need to measure orientation to both mainstream U.S. and Mexican culture and to account for SES when examining cultural orientation (Negy & Woods, 1992). Controlling for SES, we expected Mexican orientation to be associated with family-oriented patterns of expectations (e.g., delay moving out of parents’ home until marriage) and mainstream orientation to be associated with individualistic patterns of expectations (e.g., delay marriage and pursue a professional career).

Cultural values

In addition to assessing overall cultural orientation, it is important to measure specific cultural values that can guide behavior and future goals (Greenfield et al., 2003). Within-group variability in cultural values may be linked to different patterns of future expectations. In particular, familism and traditional gender role attitudes are important values that are associated with Mexican culture (Marín & Gamba, 2003). Studies suggest that gender role attitudes of immigrants and Mexican Americans are significantly more traditional than non-Hispanic white Americans (Rodriguez, Ramirez, & Korman, 1999), though they become less traditional as more time is spent in the U.S. (Valentine & Mosley, 2000). Traditional attitudes have been linked to Mexican American high school girls’ lower educational and career expectations (McWhirter, Hackett, & Bandalos, 1998), and to adolescents’ decreased expectations for establishing an independent household before marriage (Goldscheider & Goldscheider, 1992). These studies led us to expect that traditional attitudes would be associated with family-oriented patterns of future plans (e.g., expect early marriage and lower career attainment, particularly for girls) and less traditional attitudes would be associated with individualistic patterns (e.g., more time as a single head-of-household).

Familism appears to be an enduring Latino cultural value and includes beliefs about maintaining close family relationships into adulthood (Sabogal, Marín, Otero-Sabogal, Marín, & Perez-Stable, 1987). Adolescents’ familism values may be associated with their plans for young adulthood. For example, a sense of family obligations is associated with a greater likelihood of living at home and pursuing higher education after high school (Fuligni & Pedersen, 2002). Further, Mexican American adolescents who value family support tend to have fewer externalizing problems and better academic engagement (Gonzales et al., 2008). Therefore, we expected that high familism values would be linked to congruent and family-oriented patterns of future expectations (e.g., marry early and enter employment early). In contrast, adolescents with lower familism values would have less congruent and more individualistic patterns of future plans.

Cultural involvement

Ecological perspectives propose that how and with whom adolescents spend their time provides opportunities for learning skills, forming a sense of identity, and developing interpersonal bonds (Bronfenbrenner, 1979). Time spent with companions may reflect adolescents’ future plans because companions transmit cultural norms and beliefs regarding work and family roles. The family context is usually a primary source for information about successfully negotiating important developmental tasks in adulthood (e.g., Bryant, Zvonkovic, & Reynolds, 2006; Starrels & Holm, 2000). However, immigrant and ethnic minority parents’ knowledge of U.S. norms and institutions may be limited. Particularly in these cases, the peer context may play an important role in socializing future plans and goals.

To explore associations between adolescents’ cultural involvement and future plans, we studied time spent with Mexican family members and with Mexican peers. As our measure does not indicate the quality or content of time spent in particular contexts, we drew upon past research focusing on the notion of unsupervised peer time to formulate our hypotheses. For example, Latino parents report socializing their children to be successful in marriage and family and to hold high education and career aspirations (Azmitia & Brown, 2002). Yet, time spent in unsupervised peer contexts has been associated with unproductive and deviant activities, which may not be consistent with success in future adult roles (Larson & Seepersad, 2003). Therefore, we expected that more time spent with family would be associated with patterns of higher and more time spent with peers would be associated with patterns of lower educational goals and family role orientations.

Links between Patterns of Future Expectations and Adjustment

Developmental task theory suggests that adjustment patterns earlier in life are linked to adjustment patterns in adulthood (Masten, Burt, Roisman, Obradović, Long, & Tellegen, 2004). Further, the timing of adult role attainment can impact well-being over the life course. Theory indicates that transition sequences that foster gains in human, social, and personal capital are linked to better adjustment in adulthood (Caspi, Wright, Moffitt, & Silva, 1998). In particular, delaying marriage and parenthood and pursuing higher education are important determinants of job attainment and future earnings (Mouw, 2005). Adult role transitions are also linked to factors present during childhood and adolescence. For example, non-Hispanic white girls who exhibit risky behavior and problems at school move out earlier than do girls with fewer problems (Stattin & Magnusson, 1996). Therefore, understanding the extent to which adolescents’ well-being is associated with patterns of future plans increases our understanding of how pathways to adulthood are linked to adjustment over the life course.

Two indicators of individual adjustment, depressive symptoms and risky behavior, are particularly relevant given that these problems often surface for the first time during adolescence (Avenevoli & Steinberg, 2001; Moffitt, 1993). Problematic adjustment may hinder adolescents’ expectations for future success in familial, educational, and work roles. Research with Israeli (Seginer & Lilach, 2004) and Hawaiian (Skorikov, 2007) adolescents suggests that depression is linked to expectations for never marrying and unrealistic educational and career plans. In addition, adolescent risky behavior is linked to lower expectations for marriage (Manning, Longmore, & Giordano, 2007), and lower educational and occupational attainment (e.g., Tanner, Davies, O’Grady, 1999). We expected risky behavior and depression to be associated with patterns of future plans that were less congruent, where family roles were avoided (i.e., move out of parents’ home early, delay marriage), and where educational and career goals were lower.

Relationship quality with parents is another indicator of adolescent well-being. Parents are important sources of information about future adult roles (Tucker, Barber, & Eccles, 2001). Warm and close parent-adolescent relationships are linked to increased exploration toward and self-efficacy for future work and family roles in Israeli adolescents (Seginer, Vermulst, & Shoyer, 2004) and to delayed family role expectations in rural non-Hispanic white adolescents (Crouter, Carson, Vicary, & Butler, 1988). Furthermore, longitudinal data with non-Hispanic whites have documented that transitions into adult roles generally lead to more emotional closeness with parents (e.g., Aquilino, 1997). Taken together, these studies suggest that strained parent-child relationships may be a symptom or a predictor of adolescent goal-setting that does not indicate progression toward mature roles or an understanding of the steps required to reach future goals. We expected that parent-adolescent warmth would be positively associated with patterns of realistic expectations (i.e., congruent plans). Alternatively, first generation immigrant parents and their adolescents may disagree on what constitutes realistic goals due to dissonant acculturation (Portes & Rumbaut, 2001). In this case, adolescents may report realistic future plans even in the context of less warmth with their parents.

Moderating Factors

Socioeconomic status is an important contextual factor for understanding adolescents’ future expectations (García Coll et al., 1996). For example, potential negative associations between risky behavior and patterns of low educational and career goals may be stronger for lower than for higher SES adolescents because higher SES adolescents have greater access to resources that can minimize negative associations. Feminist perspectives also draw attention to gender differences in access to work and family roles (e.g., Coltrane & Parke, 1998). While most adolescents report that successful familial relationships are a high priority, girls are more likely than boys to mention occupational and familial goals at equal frequencies when asked about their futures (e.g., Yowell, 2000). Therefore, SES and gender were examined as moderators of the links between future expectations and adolescents’ cultural experiences and adjustment.

Method

Participants

Interviews were conducted with 246 Mexican American families as part of a study of family socialization and adolescent development (author citation). Families were eligible for participation if (1) 7^th graders, an older sibling, biological mothers and biological or long-term adoptive (at least 10 years) were living in the home, (2) mothers were of Mexican origin, and (3) fathers worked at least 20 hours per week. Though not a requirement, 93% of fathers were also of Mexican origin. These criteria were set given the goals of the larger study, to understand both mothers’ and fathers’ roles and how they were shaped by parent-work dynamics. We focus on 7^th graders (target adolescents in the larger study) to understand young adolescents’ future plans.

Families represented a diverse range of backgrounds. The median family income level was $40,000 (SD = $45,382, range = $3,000 – $400,000) and parents reported approximately ten years of education (M = 10.33, SD = 3.74 for mothers; and M = 9.88, SD = 4.37 for fathers). Most fathers (69%) and mothers (71%) were born in Mexico. Parents born outside of the country had lived in the U.S. an average of 12.38 years for mothers (SD = 8.86) and 15.18 years for fathers (SD = 8.78). Target adolescents (7^th graders) were 51% female (n = 125) and about 12.51 years old (SD = 0.58; range = 11 – 14). Most (62%, n = 153) were born in the U.S. and most (84%, n = 207) completed their interview in English.

Procedures

Names of seventh-grade adolescents with a Latino background who were not learning disabled were obtained from five school districts and five parochial schools in and around a large southwestern city. Letters were mailed to families of these adolescents (N = 1,856) describing the goals of the study in both English and Spanish. Follow-up phone calls were then made to assess interest in and eligibility for participation. For 438 families (24%), the contact information was incorrect and repeated attempts to find updated information through school personnel or public listings were unsuccessful. An additional 42 (2.4%) families moved between the initial screening and final recruitment contact, and 148 (8%) refused to be screened for eligibility. Eligible families included 421 families (32% of those we were able to contact and screen for eligibility). Of those who were eligible (n = 421), 284 (or 67%) agreed to participate, 95 (23%) refused, and we were unable to recontact the remaining 42 families (10%) who were eligible to determine if they would participate. Interviews were completed by 246 families. The remaining 38 families were unable to be located or refused to participate at the time of the interview.

Data were collected from mothers, fathers, and adolescents using two procedures. First, family members completed in-home interviews in their preferred language (English or Spanish) led by trained bilingual interviewers. Next, within three to four weeks of the home interview, adolescents participated in a series of seven nightly phone calls (five weekday calls and two weekend calls) about their daily time use and each parent participated in four phone calls (three weekday calls and one weekend call). Using a cued-recall procedure (McHale, Crouter, & Bartko, 1992), adolescents reported on their involvement during non-school hours in 86 different activities, including how long each event lasted and who else was present (i.e., gender, ethnicity, and relationship to adolescent). Parents provided the same types of information about activities they participated in with target adolescents and about their involvement in housework. Families received a $100 honorarium for participating in the home interviews and $100 for participating in the phone interviews.

Measures

All measures were forward and back translated into Spanish by two translators familiar with the local Spanish dialect using the method outlined by Foster and Martinez (1995).

Future expectations

Adolescents’ expectations for future role transitions and educational and occupational attainment were assessed with four items: (1) “How old do you expect to be when you move out of your parents’ home?”; (2) “What age do you expect to be when you get married?”; (3) “How far do you really think you will go in school?”; and (4) “What kind of job would you like to have when you are an adult?” Response choices for education were on a continuous scale representing the total number of years of education (e.g., 16 = bachelor’s degree; 21 = J.D., M.D., Ph.D.). Adolescents’ job preferences were coded for prestige (Nakao & Treas, 1994), and ranged from occupations such as waiter/waitress (28.08) to physician (86.05).

Family background

Parents reported their income, educational attainment, years living in the U.S., and adolescents’ country of birth. Adolescents’ language preference (English or Spanish) for the interview was also recorded. A socioeconomic status variable was computed by log transforming family income to correct for high positive skewness, and then standardizing the transformed income score, along with mothers’ and fathers’ education attainment level into z-scores. These three scores were then averaged (α = .78; SD = .84; range = −2.36 – 2.08).

Cultural experiences

Adolescents’ cultural experiences were assessed with two indicators each of cultural orientation, cultural values, and involvement. Adolescents’ cultural orientations to Mexican and mainstream U.S. culture were assessed with the Acculturation Rating Scale for Mexican Americans-II (Cuéllar et al., 1995). Adolescents responded to 30 items (e.g., “My friends are of Mexican origin” and “I enjoy English language TV”) on a 5-point scale (1 = not at all, 5 = extremely often or almost always). Scores from this measure included Mexican orientation (17 items; α = .90) and mainstream orientation (13 items; α = .82).

Two cultural values were also examined. Gender role attitudes were assessed with 10 items (e.g., A husband’s job is more important than a wife’s) on a 4-point scale (1 = strongly agree, 4 = strongly disagree; Hoffman & Kloska, 1995). Items were averaged, with higher scores indicating more traditional attitudes (α = .82). Familism support (α = .74) was measured with a subscale from the Mexican American Cultural Values Scale (Knight et al., under review). Adolescents rated six items (e.g., Parents should teach their children that the family always comes first) on a 5-point scale (1 = strongly disagree, 5 = strongly agree).

The third aspect of cultural experiences was the amount of time that adolescents spent in particular cultural contexts. Using data from the phone interviews, variables were computed for time spent with family members (only nuclear or extended kin present) and Mexican peers (only non-sibling kin and non-kin under age 18 present). For time spent with family members, families in which the father was not of Mexican origin (n = 18, as well as one family who was missing this information) were excluded to increase the likelihood that the variable represented time spent with Mexican family members. Time use variables were converted to proportion scores, computed as time in the selected context divided by total time reported across the seven phone calls, and square root transformed to correct for skewness. In analyses involving time use variables, untransformed overall means (in hours) are reported for ease of interpretation, though effect sizes were computed using the transformed proportion scores. Adolescents’ reports of their time use were highly reliable; correlations between younger siblings’ reports and reports of the other target family members ranged from .80 to .90 for shared time.

Adjustment

Adolescents’ well-being was assessed by depressive symptoms, risky behaviors, and parent-adolescent warmth. Adolescents reported how often they experienced depressive symptoms (e.g., I thought my life had been a failure) on a 4-point scale (1 = rarely or none of the time, 4 = most of the time) using the 20-item Center for Epidemiological Studies Depression Scale (α = .85; Radloff, 1977). The CES-D has established measurement equivalence for Mexican American adolescents (Crockett, Randall, Shen, Russell, & Driscoll, 2005). Adolescents reported how often they engaged in each of 24 risky behaviors (e.g., smoked cigarettes, started a fight with someone) on a 4-point scale (1 = never, 4 = more than 10 times) using a measure adapted from Eccles and Barber (1990; α = .91) for use with ethnically diverse youth. Finally, parent-adolescent warmth was assessed with the acceptance subscale of the Child Report of Parental Behavior Inventory (Schaefer, 1965). Adolescents responded to eight items (e.g., My mother/father is able to make me feel better when I am upset) on a 5-point scale (1 = almost never, 5 = almost always) for their mothers (α = .84) and fathers (α = .89) separately. Cross ethnic and language equivalence of CRPBI acceptance scale has been established by Knight and colleagues (Knight, Tein, & Shell, 1992; Knight, Virdin, & Roosa, 1994).

Results

Results are organized around the three goals of the study: (a) to describe patterns of Mexican American adolescents’ expectations for future work and family roles, and to examine links between patterns of future expectations and (b) adolescents’ cultural experiences and (c) adolescents’ adjustment. Our plan of analysis included a clustering strategy and then a series of ANCOVAs. First, a cluster analysis was conducted using Sleipner version 2.1 (Bergman & El-Khouri, 2002) to identify unique patterns of adolescents’ future expectations. Next, we tested a series of 5 (Cluster) × 2 (Gender) ANCOVA models, with familial SES included as a covariate, to determine whether future expectation clusters differed significantly by cultural experiences and adjustment. We also tested a series of 5 (Cluster) × 2 (Gender) × 2 (Socioeconomic Status) ANOVA models, in which SES was divided into two groups according to a median split, to determine whether SES moderated associations between future expectations and the culture variables. Follow-up comparisons were conducted with Tukey’s tests using unadjusted means.

Goal 1: Identify Patterns of Mexican American Adolescents’ Future Expectations

Descriptive statistics and correlations (Table 1) suggest that adolescents’ future plans are linked in expected ways. On average, adolescents who plan to move out of their parents’ house later also expect to marry later. In turn, expecting to marry later is linked to higher educational expectations, and higher educational expectations are linked to higher career goals.

Table 1.

Correlations and Descriptive Information for Adolescents’ Future Expectations, Cultural Experiences, and Adjustment

	Descriptives		Correlations
	M	SD	1	2	3	4
1. Expected age to move	19.92	2.59	–
2. Expected age to marry	23.87	2.90	.20*	–
3. Educational expectation	15.66	2.17	.01	.14*	–
4. Career aspirationa	61.66	14.61	−.01	.02	.26*	–
5. Mexican orientation	3.75	0.72	.28*	−.04	−.04	−.11
6. Mainstream orientation	3.95	0.60	−.15*	.00	.07	.08
7. Gender role attitudesb	2.23	0.49	.14*	−.14*	−.24*	−.04
8. Familism support	4.39	0.58	.04	.16*	.19*	−.01
9. Mexican family timec,d	24.96	10.37	.19*	−.09	−.01	−.06
10. Mexican peer timec	3.09	3.46	−.03	−.08	−.30*	−.17*
11. Depressive symptoms	1.82	0.49	−.00	−.05	−.22*	−.09
12. Risky behavior	1.37	0.40	−.10	.10	−.27*	−.13*
13. Parental warmthe	3.88	0.63	.17*	.03	.31*	−.02

Note.

^aHigher values correspond to more prestigious career aspirations.

^bHigher values indicate more traditional gender role attitudes.

^cUntransformed overall means (in hours) across one week are reported.

^dAdolescents whose father was not of Mexican origin were not included.

^eAverage of mother warmth and father warmth.

^*p< .05.

^**p < .01.

Adolescents with complete data on the four future expectation variables (n = 237; 96% of the sample) were entered into a cluster analysis to identify unique patterns of expectations. Adolescents with missing data were not significantly different from the rest of the sample on family background characteristics or the dependent variables. Cluster analysis is sensitive to outliers (Bergman, 1988), and nine cases were removed by the standard procedure in Sleipner (i.e., a case is retained if the squared Euclidean distance from its closest neighbor is .50 or less).¹

The cases that were retained (n = 228; 93% of the sample) were then clustered according to the weighted average linkage algorithm using squared Euclidean distance. The best solution was chosen based on the merging coefficients, explained error sums of squares (ESS), and whether solutions distinguished among theoretically meaningful groups. Similar to interpreting scree plots in exploratory factor analysis, it is recommended to select the cluster solution just before a large increase in the merging coefficient and before a large decrease in the explained ESS (Aldenderfer & Blashfield, 1984). For the 7, 6, 5, 4 and 3 cluster solutions, respectively, the merging coefficients were 1.99, 2.00, 2.41, 2.75, and 3.46 and the explained ESS were 53.56, 49.13, 47.98, 40.75, and 33.07. The merging coefficients supported the 6 cluster solution and the explained ESS supported the 5 cluster solution. Closer inspection revealed that the 6 cluster solution included two small clusters (n = 16 and n = 5) that were theoretically similar. We determined that the 5 cluster solution optimized theoretical relevance and the cluster statistics.

Next, to increase within-cluster homogeneity, cases were relocated using an iterative procedure (Aldenderfer & Blashfield, 1984). A total of 65 cases were relocated, resulting in an increase in the explained ESS from 47.98% to 54.70%. The means of each variable within each of the five clusters are shown in Table 2. The first cluster (Career Oriented) had the highest educational and career goals. Adolescents in the second cluster (Independent) expected the largest gap in time between age to move out of parents’ house and age to get married (7.2 years). In contrast to the Independent cluster, adolescents in the third cluster (Family Oriented) had the smallest gap in time between expected ages to move out and to marry (0.8 years). Adolescents in cluster four (Early) were characterized by the earliest expected family role transitions and the lowest school and career goals. Finally, adolescents in cluster five (Inconsistent) expected to move out and to get married at an earlier than average age, but had higher than average career goals and lower than average educational goals. Their career aspirations seemed inconsistent with their expectations for (relatively) low educational attainment and early transitions into family roles².

Table 2.

Means (and Standard Deviations) of Adolescents’ Future Expectations by Cluster and for the Full Sample

	Career Oriented	Independent	Family Oriented	Early	Inconsistent	Full Sample
Age to move out	19.36a (1.11)	18.90c (1.15)	24.21abcd (1.41)	19.34b (1.43)	18.93d (1.20)	19.62 (2.01)
Age to get married	23.67a (2.34)	26.10ac (2.07)	25.00b (1.77)	22.64bc (2.30)	21.17ab (1.40)	23.69 (2.77)
Educational attainment	18.88ab (1.71)	15.75bc (1.26)	15.79a (1.64)	14.57ac (1.97)	14.71b (1.88)	15.71 (2.18)
Career aspiration	76.27abc (10.55)	63.58bd (8.48)	68.42a (8.27)	41.31ade (6.43)	67.06ce (7.40)	62.53 (13.77)
N	33	68	24	44	59	228

Note. Clusters with the same subscripts in each row are significantly different at p < .01 according to Tukey’s post hoc tests.

Background characteristics of each cluster are presented in Table 3. Compared to the other clusters, Career Oriented adolescents were of higher SES and Family Oriented adolescents were of a lower generation status. In addition, compared to the other clusters, a higher percentage of adolescents in the Family Oriented and Early clusters preferred Spanish and were born in Mexico, Career Oriented adolescents were predominately female, and boys were overrepresented in the Independent and Inconsistent clusters.

Table 3.

Cluster Group Differences in Family and Cultural Background

	Career Oriented	Independent	Family Oriented	Early	Inconsistent
Background Variables	M (SD)	M (SD)	M (SD)	M (SD)	M (SD)	FCluster (df)
Socioeconomic statusa	0.42ab (0.88)	0.10 (0.76)	−0.32b (0.88)	−0.17a (0.83)	−0.03 (0.83)	3.69** (4,223)
Mothers’ years living in the U.S.	23.06a (15.01)	23.79b (14.20)	11.46abc (11.69)	17.07 (14.53)	23.47c (15.53)	4.63* (4,223)
Adolescents’ generation	2.16a (1.17)	2.41b (1.08)	1.28abc (0.58)	2.00 (1.03)	2.32c (1.10)	5.51** (4,215)
	N	N	N	N	N	χ2 (4)
Country of birth
Mexico	12	18	18	17	16	20.78**
U.S.	21	50	6	27	43
Language preference
Spanish	3	4	9	12	6	21.25**
English	30	64	15	32	53
Gender
Girls	27	30	14	25	25	16.37**
Boys	6	38	10	19	34

Note. Clusters with the same subscripts in each row are significantly different at p < .05 according to Tukey’s post hoc tests.

^aSocioeconomic status is a standardized variable and is the composite of family income, mothers’ education attainment, and fathers’ education attainment.

^*p< .05.

^**p < .01.

Goal 2: Links between Patterns of Future Expectations and Cultural Experiences

Three sets of indicators were tested to assess whether patterns of future expectations were associated with different aspects of cultural experiences: (1) orientation to Mexican culture and mainstream culture, (2) endorsement of traditional gender role attitudes and familism, and (3) time spent in Mexican familial and peer contexts. Because analyses revealed that SES did not significantly moderate associations between future expectations and cultural experiences³, SES was added as a covariate. Mean values of the three indicators of cultural experience by the clusters are presented in Table 4. As shown in Table 5, the ANCOVA models revealed that gender or the interaction between gender and cluster did not predict meaningful differences in cultural experiences. However, there were systematic differences in the three sets of indicators of cultural experiences across the clusters.

Table 4.

Cluster Means (and Standard Deviations) on Indicators of Cultural Experiences and Adjustment

	Career Oriented	Independent	Family Oriented	Early	Inconsistent
Cultural orientations
Mexican orientation	3.55 (0.70)	3.45 (0.74)	4.12 (0.67)	3.90 (0.75)	3.50 (0.80)
Mainstream orientation	4.06 (0.45)	4.01 (0.51)	3.70 (0.76)	3.86 (0.68)	4.11 (0.57)
Differencea	−0.51a (1.01)	−0.57bd (1.05)	0.42abc (1.13)	0.04de (1.08)	−0.61ce (1.13)
Cultural values
Gender role att.b	2.03a (0.52)	2.06b (0.52)	2.33 (0.51)	2.27 (0.47)	2.37ab (0.42)
Familism support	4.56a (0.35)	4.44 (0.49)	4.53 (0.72)	4.44 (0.46)	4.20a (0.72)
Cultural involvementc
with family membersd	24.88 (10.84)	23.06a (10.09)	26.08 (8.57)	27.49ab (10.11)	23.93b (11.39)
with Mexican peers	1.69ab (3.18)	2.62 (2.73)	2.20 (2.36)	3.03a (3.22)	4.01b (4.42)
Adjustment
Depressive symptoms	1.68 (0.46)	1.73 (0.48)	1.83 (0.36)	1.94 (0.58)	1.89 (0.46)
Risky behavior	1.18ab (0.17)	1.41a (0.39)	1.24 (0.20)	1.32 (0.30)	1.46b (0.47)
Parental warmthe	4.11a (0.62)	3.80b (0.62)	4.23bc (0.49)	3.82 (0.56)	3.70ac (0.69)

Note. Clusters with the same subscripts in each row are significantly different at p < .05 according to Tukey’s post hoc tests.

^aDifference scores were created by subtracting Mainstream Orientation from Mexican Orientation.

^bHigher values correspond to more traditional gender role attitudes.

^cUntransformed overall means (in hours) across one week are reported.

^dAdolescents whose father was not of Mexican origin were excluded.

^eParental warmth is the average of mother warmth and father warmth.

Table 5.

Analysis of Covariance Results for Cluster Differences in Cultural Experiences and Adjustment

	Overall F	SES	Cluster	Gender	Cluster × Gender
Cultural orientationsa
Mexican vs. Mainstream	10.20**	9.70**	.75	3.22	.35
Cultural values
Gender role attitudesb	6.44**	26.92**	4.01**	0.72	2.21†
Familism support	2.00*	5.67*	2.46*	0.02	0.56
Cultural involvement
with family membersc	2.56**	6.04**	3.63**	0.31	0.73
with Mexican peers	2.50**	5.16*	3.51**	0.12	0.67
Adjustment
Depressive symptoms	2.21*	8.43**	1.44	2.94	0.04
Risky behavior	2.23*	0.01	3.13*	0.73	0.95
Parental warmthd	2.49**	5.94*	4.23**	0.08	0.04

Note.

^aOverall F column is the Cluster × Dimension F value for cultural orientations.

^bHigher values correspond to more traditional gender role attitudes.

^cAdolescents whose father was not of Mexican origin were not included.

^dWarmth with mothers and fathers were analyzed as a within-groups factor, and findings revealed cluster group differences in the average of mothers’ and fathers’ reports of warmth. The Cluster Group × Parent (Mother vs. Father) interaction was not significant.

^*p< .05.

^**p < .01.

Mexican and mainstream cultural orientations were included in the same analysis with cultural orientation treated as a within-group factor to test cluster membership differences in both Mexican and mainstream cultural orientations simultaneously. The follow-up analyses of the significant cluster by cultural orientation interaction (comparing the difference scores of Mexican orientation minus mainstream orientation across groups) revealed that Family Oriented youth reported stronger Mexican than mainstream orientations and differed significantly from Career Oriented, Inconsistent, and Independent youth, who reported stronger mainstream than Mexican orientations (d’s = .84, .91 and .95, respectively). In addition, Inconsistent and Independent youth also differed from Early youth (d’s = .51 and .50, respectively): Inconsistent and Independent youth reported stronger orientations toward mainstream culture as compared to Mexican culture, but Early youth had similar orientations to Mexican and mainstream cultures.

For cultural values, there were significant overall cluster effects for both traditional gender role attitudes and familism values. Inconsistent adolescents reported gender role attitudes that were significantly more traditional compared to Career Oriented or Independent adolescents (d’s = .72 and .66, respectively). Furthermore, Inconsistent adolescents reported lower familism values than Career Oriented adolescents (d = .64).

Models examining adolescents’ cultural involvement revealed significant overall cluster effects for time spent with family and Mexican peers. Early adolescents spent more time with family than Independent or Inconsistent adolescents (d’s = .44 and .33, respectively). In addition, Early and Inconsistent adolescents spent more time with Mexican peers compared to Career Oriented adolescents (d’s = .62 and .68, respectively).

Goal 3: Links between Patterns of Future Expectations and Adjustment

Cluster means of depressive symptoms, risky behaviors, and average warmth with parents are shown in Table 4. Analyses did not support SES as a moderator between future plans and adjustment.⁴ Results of the models controlling for SES are presented in Table 5. There were no significant overall gender effects or cluster by gender interactions in these models. The overall cluster effect was not significant for depressive symptoms, but was for risky behavior and parental warmth. Follow-ups revealed that adolescents in the Career Oriented cluster reported significantly less risky behavior compared to adolescents in the Independent and Inconsistent clusters (d’s = .76 and .79, respectively) and significantly more warmth with their parents compared to adolescents in the Inconsistent cluster (d’s = .62). Adolescents in the Family Oriented cluster also reported significantly more warmth with their parents compared to adolescents in the Independent or Inconsistent clusters (d = .77 and .88, respectively).

Discussion

The process of transitioning from adolescence to adulthood depends on access to resources and cultural norms and beliefs about when and how transitions should occur. This study provides a “snapshot” of this process by describing how Mexican American 7^th graders think about their work and family futures. We also gain understanding of how context relates to this process by describing links between patterns of future plans and cultural and adjustment factors. In discussing our results, we reflect upon the cultural-ecological context to understand how experiences, future plans, and pathways are linked over the life course.

Our pattern-analytic approach identified groups of adolescents characterized by distinct patterns of future plans. For example, the cluster strategy showed that both Independent and Family Oriented adolescents plan to delay marriage, but that only Family Oriented youth plan to live with their parents in the years before marriage. This difference in marriage expectations would be obscured had we only examined bivariate correlations between future plans, and underscores the importance of understanding patterns across key decisions about adult roles. Examination of the collective set of expectations also revealed patterns of developmental significance. Over a quarter of the sample fell into the Inconsistent cluster, which was characterized by low levels of expected educational attainment coupled with high career expectations. In early adolescence, youth may have unrealistic expectations and a less clear understanding of what is necessary to obtain particular adult careers and lifestyles. It is likely that the patterns of adolescents’ expectations for their adult lives will shift over the course of adolescence as they develop more sophisticated cognitive abilities and acquire experiences in their journeys toward adulthood.

Important cluster differences emerged with regard to access to resources (SES) and in cultural background characteristics that reflect contact with the U.S. (i.e., adolescents’ language preference, nativity, generation status, and mothers’ years living in the U.S.). Although it is normative for non-Hispanic whites to move out of parents’ home several years prior to getting married (Cohen et al., 2003), this study indicates that Mexican American youth who are born in Mexico and have spent less time living in the U.S. (i.e., Family Oriented cluster) are less likely to expect this path. Socioeconomic status was lowest in the Early and Family Oriented clusters. For the Early cluster, being U.S. born and speaking English was linked to lower educational and career goals and for the Family Oriented cluster, less contact with the U.S. (in terms of generation status, mothers’ years living in the U.S., and youth’s nativity) was linked to higher educational and career goals. In contrast, the cluster with the highest SES and the longest number of years living in the U.S. reported the highest educational and career goals (i.e., Career Oriented cluster). These adolescents may have accrued more resources in terms of both financial resources and cultural experiences that provide a context for expectations to pursue extended education and professional careers.

Results from the second goal of the study revealed that adolescents’ future expectations are linked to their cultural experiences in complex ways. An important contribution of this study was to move beyond the focus on singular and proxy measures of culture (e.g., generation status or nativity) to understand how multiple dimensions of culture (i.e., cultural orientations, values, and involvement) are linked to youth’s future plans. Such an approach allows for a more nuanced perspective on the role of culture in Mexican American adolescents’ future plans. For instance, endorsing less traditional gender role attitudes is consistent with Career Oriented adolescents’ (82% female) high education and career goals and Independent adolescents’ plans to delay marriage. Family Oriented adolescents who expected to spend more time living with parents and aspired to a more professional career also reported higher Mexican than mainstream orientations. Most important may be what we learned by looking across the constellation of cultural experiences. In the Inconsistent cluster, for example, adolescents may receive contradictory messages about adult roles. In particular, their traditional gender role attitudes seem to conflict with their low familism values and higher mainstream as compared to Mexican orientations. Further, Inconsistent adolescents spent the most time in unsupervised Mexican peer contexts. Perhaps protective cultural values, such as familism, are not advanced in the peer context to the same extent as in the familial context. Taken together, our findings suggest that young adolescents’ cultural experiences are a factor in their expectations for adulthood and that the complexities involved in studying the role of culture is evident when culture is conceptualized in a multidimensional way.

Clusters did not differ by depressive symptoms, but higher parental warmth was linked to cluster membership emphasizing either future work roles (i.e., Career Oriented adolescents) or family roles (i.e., Family Oriented adolescents). In addition, Career Oriented youth reported significantly less involvement in risky behaviors than Independent and Inconsistent adolescents. One potential explanation is that high education and career goals, like those of Career Oriented adolescents, are protective in terms of involvement in risky behavior. In addition, Career Oriented adolescents spend less time with Mexican peers and describe higher levels of parental warmth than Inconsistent adolescents, both of which may be additional protective factors for involvement in risky behaviors (e.g., Osgood, Anderson, & Shaffer, 2005). The pattern for the Independent cluster suggests that adolescents who engage in risky activities plan to delay marriage, perhaps because they expect to continue engaging in risky activities in young adulthood.

We tested whether gender moderated associations between cluster membership and cultural experiences and well-being, but no interactions were significant. For Mexican Americans in early adolescence, it seems that boys’ and girls’ plans and goals for their future are associated with culture and well-being in similar ways. Gender may begin to differentiate links between future expectations and cultural experiences and adjustment as adolescents mature physically and socially.

SES did not emerge as a moderator of associations either. Past research has shown that SES is linked to adolescents’ future education goals and family plans (e.g., East, 1998). Therefore, it is not surprising that SES was intimately tied to the clusters that emerged, and may explain why cluster × SES interactions were not significant. In addition, the confounds among indices of culture and SES (McLoyd, 1998; Negy & Woods, 1992) suggest that future work with larger samples is needed to begin to understand how culture and SES separately and in combination are linked to girls’ versus boys’ future plans.

Limitations and Future Research Directions

The limitations of the current study offer a springboard for future research. First, the exploratory nature of cluster analysis makes replication particularly important. Though our sample is diverse in terms of family and cultural background, its representativeness is limited to young Mexican American adolescents living with both parents and at least one other sibling in a large Southwestern city. In addition to increasing generalizability, future research should include measures of other future plans (e.g., cohabitation) in tracing pathways to adulthood.

A second limitation is our focus on future plans and goals. In reality, young people are not always able to realize their goals or follow through on their plans, particularly in the face of contextual constraints. Discrimination, poverty, and language barriers can hinder educational and career progress, and the availability of a partner has obvious consequences for the timing of marriage and parenthood. An important step for future research will be to understand how much personal agency and decision-making play a role in the transition to adulthood versus contextual resources and constraints (and fortuitous life events).

Finally, the cross-sectional nature of the data does not allow us to draw conclusions regarding direction of effects. It is reasonable that cultural orientation and values lead to expectations about adult roles; however, adolescents also may adapt their cultural involvement to match that of their future expectations. For example, adolescents who are particularly motivated to achieve educational and career success might adopt cultural values of competitiveness and independence that will facilitate their progress. Longitudinal studies are needed to uncover the ways that culture, adjustment, and pathways to adulthood are related.

In conclusion, the present study offers evidence for diversity in early Mexican American adolescents’ future expectations. It appears that there are distinct groups of adolescents who vary widely in their expected pathways to adulthood. Further, while all elements of culture are not linked in the same ways to youth’s patterns of future expectations, there is evidence that these expectations are associated with adolescents’ culture and well-being in meaningful, albeit complex, ways. As the growing population of Mexican American adolescents living in the U.S. strive to reach their work and family goals, research on their goals for adulthood will become increasingly important.

Biographies

Emily Cansler earned her M.S. in Family and Human Development at Arizona State University. She is now a Survey Coordinator in the Survey Research Group at the RAND Corporation. Her research interests include culture, the transition to young adulthood, and adjustment in Mexican American youth.

Kimberly A. Updegraff received her Ph.D. in Human Development and Family Studies from Pennsylvania State University. She is a Professor in the School of Social and Family Dynamics at Arizona State University. Her research focuses on culture and gender socialization processes in family and peer relationships.

Sandra D. Simpkins received her Ph.D. in Developmental Psychology from UC-Riverside. She is an Assistant Professor in the School of Social and Family Dynamics at Arizona State University. Her research focuses on children’s out-of-school activities/programs, friendships, the family context, and the impact of these phenomena on developmental outcomes