1 Introduction

1.1 Subjective Well-Being and Affect

Living the good life has come to be identified by psychological researchers as subjective well-being (SWB), or psychological well-being (Diener 2000). Life satisfaction, or a person’s assessment of the quality of her or his life (Diener 1994), represents the cognitive judgment component of SWB. According to Lucas et al. (1996), when making an evaluation of life satisfaction, the person examines the physical aspects of her or his life, weighs the good against the bad, and then comes to an overall conclusion of their satisfaction.

The other component of SWB is affect (positive and negative; Diener 1994). Positive affect includes positive emotional states, such as joy, interest, and alertness, whereas negative affect includes emotional states, such as sadness, fear, and anger (Watson 2002). The initial distinction between positive and negative affect was discussed in Bradburn’s (1969) pioneering work. His research focused on how macro-level changes, such as changes in education levels, employment patterns, and urbanization, affected individuals’ life situations, and in turn, their happiness or well-being. Bradburn operationalized happiness as the balance between positive and negative affect, based on his investigations revealing that positive and negative affect were distinct dimensions of well-being. Consistent with Bradburn’s findings, Diener (2000) found that happy persons reported mild to moderately pleasant emotions most of the time, and in different situations, including when alone, with others, and when working or at leisure. Therefore, positive affect, more so than negative affect, appears to be strongly related to well-being.

Few studies have analyzed the distinction between the affective components of SWB and the cognitive judgment of life satisfaction. Lucas et al. (1996) examined the discriminant validity of life satisfaction, positive affect, negative affect, self-esteem, and optimism using multitrait–multimethod matrix analysis on three sets of data with college students. Results demonstrated that life satisfaction was related to, but discriminable from, positive and negative affect, and that positive affect was discriminable from negative affect. Furthermore, Diener et al. (1999) suggested that life satisfaction, positive affect, and negative affect be assessed separately in order to better understand the overall components of SWB. In the present study, the life satisfaction and positive affect of Latinas are examined and measured as two separate components of SWB.

1.2 Subjective Well-Being and Latinos

Another area of SWB research includes the examination of cultural differences. Triandis (2000) state that societies have different sets of values; therefore, individuals are likely to consider different criteria when judging their success. Individualism is found in societies in which “the self is regarded as autonomous, personal goals have priority over in-group goals… and social exchanges characterize interpersonal relationships” (Triandis 2000, p. 13–14). On the other hand, collectivism is found in societies where “the self is regarded as an aspect of groups, interdependent with members of those groups, in-group norms have priority over personal needs… and relationships are communal” (Triandis 2000, p. 14).

Suh et al. (1998) found that persons from individualistic and collectivist cultures relied upon different criteria, particularly when determining their life satisfaction. Persons from individualistic cultures tend to rely on their affect when deciding how satisfied they are with their lives (Suh et al. 1998). Feeling pleasant emotions often is a fairly good predictor of life satisfaction for individuals in these cultures. In contrast, in collectivistic cultures, behaving in accordance with societal norms predicted higher life satisfaction. Thus, culture can significantly influence individual’s values and beliefs, which influences the factors they consider important to life satisfaction (Diener 2000).

Seligman (2002) argues that psychology must focus on and bring strength building to the forefront in the treatment and prevention of mental illness. This would be especially true for groups that have historically experienced adversity, such as ethnic minority groups. Among ethnic minority groups in the U.S., Latinos are the fastest growing group (U.S. Department of Commerce, Bureau of the Census 2010) and account for 16 % (50.5 million) of the U.S. population. California, in particular, accounted for 14 million (28 %) of the total Latino population, and people of Mexican origin comprised the largest Latino group (63 %; U.S. Department of Commerce, Bureau of the Census 2010).

A large percentage of Latinos are of lower socioeconomic class, regardless of generational status in the U.S. (Padilla et al. 1987). For example, 25.3 % of Latinos lived in poverty in 2009 compared to 9.4 % of non-Latino Whites (U.S. Department of Commerce, Bureau of the Census 2010). Research findings also suggest that disproportionally more Mexican-origin women report mental health concerns than Mexican-origin men (Espin 1997; Gil and Vazquez 1996). According to Anzaldua (1987/1999), Latinas may have poorer mental health due to their triple minority status as poor, female, and women of color. Also, Latinas are often marginalized in different contexts, including within the family, work, and education arena, which can affect their well-being. Like others living in more than one culture, Latinas also receive multiple, and often opposing, messages (Anzaldua 1987/1999), such as being told at home to indirectly communicate with others and to depend on family for social support but being told at school and work to be assertive and independent. Despite the oppressive experiences that many Mexican American women experience (Chacon et al. 1985), many of these women are also resilient. This resiliency could be due to cultural values that may buffer them from negative experiences. Some of these cultural values include familism (in the form of social support; Almeida et al. 2009), ethnic identity (Vera et al. 2011), and gender roles (Gil and Vazquez 1996). As Seligman (2002) suggests, focusing on strengths instead of weaknesses can be more effective in helping individuals, specifically the growing population of Latina women. Yet, SWB has not been examined extensively among Latinas, especially with respect to cultural factors that are relevant to this population.

As stated above, one protective factor among Latinas is the core value of familism or familismo, which is defined as Latinos’ strong loyalty, reciprocity, solidarity, and attachment to the nuclear and extended family (Marin and Marin 1991). The importance of familism to Latinos and its relationship to well-being has been well documented (Almeida et al. 2009; Rodriguez et al. 2007; Sabogal et al. 1987). However, familism has been operationalized inconsistently in the literature (Rodriguez et al. 2007). Rodriguez et al. found operational definitions ranging from a global sense of familism to subcomponents, such as family networks and support, family satisfaction, family importance, and family functioning. In addition to examining acculturation, ethnic identity, and gender role identity on SWB, the present study also examines the role of family social support on SWB, which is one aspect of familism that is important to Latinas.

1.3 Theoretical Framework: Multiracial Feminism

Multiracial Feminist Theory will be used to frame the present study and explain the relationships between the variables of acculturation, ethnic identity, gender roles, and perceived social support from family on Mexican American women’s well-being. Multiracial Feminist Theory has its origins in Feminist Theory, which was developed mainly from the work of female clinicians who came together and critiqued gender relations in the workplace and society in general (Brown 2010). Women began to join feminist consciousness-raising groups and to tell their own personal experiences of discrimination. In addition, several important documents were published that discussed sex role stereotyping, critiques on research practices and conclusions about women, and the way psychotherapy often times harmed the very women that came to seek help (Brown 2010). Early conversations of sexism were important for the development of Feminist Theory; however, women of color were often left out of these discussions and analysis. For instance, Bing and Reid (1996) argued that although feminist ideology opposes patriarchy and social oppression, it does not necessarily address and oppose racist and classist oppression.

Due to the exclusionary practices of many White feminists, multiracial feminism was developed as an evolving body of theory and practices guided primarily by women of color (Arellano and Ayala-Alcantar 2004; Baca Zinn and Dill 1996). Unlike Feminist Theory, multiracial feminism examines gender experiences interwoven with additional social factors such as race, class, ethnicity, and gender identity (Arellano and Ayala-Alcantar 2004; Baca Zinn and Dill 1996). Additionally, multiracial feminism acknowledges that these factors of race, class, ethnicity, and gender identity interact with the different settings in people’s lives and, therefore, create different experiences of power (Arellano and Ayala-Alcantar 2004; Baca Zinn and Dill 1996).

Multiracial feminism and subjective well-being are not commonly found together within the literature. Yet, they can both be understood as part of positive psychology. Positive psychology focuses on the strengths, not weaknesses, of individuals (Seligman 2002). Multiracial feminism and subjective well-being not only focus on strengths and healthy functioning of individuals (Arellano and Ayala-Alcantar 2004; Baca Zinn and Dill 1996; Comas-Diaz 1991; Diener 1984, 2000; Ryff 1989), but they also value diverse meanings of what constitutes health and the good life. As stated earlier, multiracial feminism focuses on understanding individuals based upon their unique statuses, and avoids pathologizing individuals, regardless of how different their values and behaviors are from mainstream White American culture (Comas-Diaz 1991). Subjective well-being examines cognitive and emotional states, and interpersonal frameworks that can be used to understand and foster healthy functioning among diverse individuals (Diener 2000; Diener et al. 1999). As part of positive psychology, both multiracial feminism and subjective well-being call for researchers to examine diverse constructions of life meaning, to observe how cultural experiences might shape the meaning of the good life, and to allow for individuals to clarify what works in their lives (Lopez et al. 2002).

1.4 Acculturation

Acculturation can be defined as cultural changes that occur because individuals from two distinct cultures come into continuous, first-hand contact with one another (Redfield et al. 1936). If a culture changes, the individuals of that culture change in some way as well (Cuéllar 2000). Berry’s (1980) acculturation model discusses psychological adaptation based upon the degree to which individuals adhere to either or both dominant culture and their culture of origin. Berry (1980) identified four modes of acculturation within his model. These include assimilation, integration, separation, and marginalization. Assimilation occurs when minority group members acquire the behaviors and values of the dominant culture and abandon or do not acquire, practice, or value their traditional culture. Integration occurs when minority members integrate both their traditional culture with acquired characteristics of the dominant culture. Individuals, therefore, integrate both in such a way that their ethnic identity is maintained while endorsing mainstream values as well. Separation occurs when the traditional culture is adhered to and not abandoned. Individuals’ traditional identity is also maintained, along with a reluctance to accept, change, adapt, or even identify with the host culture in which they reside or work. Lastly, marginalization occurs when minority group members do not maintain allegiance to traditional belief, values, and behaviors; individuals do not adopt the values of the dominant culture (Berry 1980). In other words, these individuals do not have a good sense of identity with either their traditional culture or the mainstream culture.

Lang et al. (1982) investigated acculturation and SWB among Latinos and found that the best-adjusted Latinos were bicultural (i.e., integrated); specifically, well-adjusted Latinos when compared to poorly adjusted Latinos had higher levels of acculturation. Guinn and Vincent (2011) also examined acculturation and life satisfaction among Mexican American women living in South Texas. However, they found that acculturation did not affect life satisfaction. The authors posit that the Mexican American women studied may “find it more compatible to remain immersed in beliefs, practices, and values of traditional Mexican culture over those of the host culture, therefore, negating any acculturation association with life satisfaction” (p. 20). Additionally, Cuéllar et al. (2004) compared acculturation and well-being of Mexican and Mexican American adults, ages 45 and older. Well-being was operationalized as a composite of life satisfaction, daily stressors, depressive symptoms, health, and self-esteem. Mexican individuals reported significantly more stress than Mexican American individuals. However, there were no differences on well-being scores between the groups (Cuéllar et al. 2004). Also, Cuéllar et al. found that acculturation significantly predicted depressive symptoms and health but not life satisfaction or self-esteem. The research discussed above indicates that findings on acculturation and its effect on SWB are inconclusive. This is likely due to studies differing in measures used to assess acculturation, differences in their target populations, and differences in definitions and measures of subjective well-being.

1.5 Ethnic Identity

Phinney (1992) defined ethnic identity as the aspect of individuals’ social identity that is derived from membership in their ethnic group and the significance individuals attach to that membership. In the U.S., members of the majority group often hold members of ethnic minority groups in low regard (Molix and Bettencourt 2010). Therefore, ethnic minorities are often devalued in the larger society and often experience a lack of power, stigmatization and discrimination, which in turn, affects their well-being (Molix and Bettencourt 2010). Branscombe et al. (1999) posited that group identification is a coping strategy that can buffer the negative effects associated with being an ethnic minority in the U.S. For example, Vera et al. (2011) explored whether ethnic identity moderated the effects of perceived discrimination on life satisfaction and positive affect among 157 early adolescent students, of which 63 % were Latino (Latino subgroups were not specified). Results indicated ethnic identity moderated the relationship between perceived discrimination and life satisfaction. However, Vera et al. (2011) also found that ethnic identity did not moderate the relationship between perceived discrimination and positive affect.

1.6 Gender Roles

Traditional gender roles play a critical role in understanding Latino well-being. Machismo is a term used to describe the construct of male gender role, which characterizes male beliefs, attitudes, and behaviors (Galanti 2003). Typically, males are expected to be strong, dominant providers who may also be hyper-masculine, aggressive and sexist (Galanti 2003). The construct marianismo, is used to describe Latina female beliefs, attitudes, and behaviors (Stevens 1973). Typically, females are expected to be nurturing, self-sacrificing, and emulate the Virgin Mary (Stevens 1973). Stevens (1973) coined the term marianismo (marianism), which springs from the religious movement of Mariology. Mariology provided a central figure, the Virgin Mary (Our Lady of Guadalupe), who is honored as the ideal woman and mother. Latin American society of all classes has accepted this description of the ideal woman (Stevens 1973). Recently, Castillo et al. (2010) developed and validated a measure of marianismo beliefs and found the measure to be consistent with the previous literature on marianismo (Stevens 1973). That is, the subscale scores on Castillo et al.’s measure of marianismo were strongly related to cognitive enculturation (i.e., learning about the Latino traditional values), interdependence, and self-sacrifice.

Although there are Latina women who deviate from traditional gender roles, many Latinas continue to feel pressure to conform to these gender roles (Stevens 1973). Furthermore, Diaz-Guerrero (1975) found that among 294 adults from Mexico City, 70 % of the sample adhered to rigid socio-cultural norms, such as masculine and feminine gender roles, that produced conflict and frustration, and these behaviors appeared to interfere with mental health. Marianismo has positive aspects that include features of loyalty, compassion, and generosity (Gil and Vazquez 1996), but the undesirable side can be responsible for conflict and confusion about the self for Latinas.

Gender roles among Latina women have been studied extensively, but mainly in relationship to traditional roles, psychological adjustment, and mental illness (Golding 1990; Napholz 1994; Vazquez-Nuttall et al. 1987). Few studies have focused on gender roles and subjective well-being among the U.S. population, and even fewer studies have explored these variables among Latina women. Bem (1975) posited that masculinity and femininity were traditionally conceptualized as bipolar ends of a single continuum. In other words, an individual had to be either masculine or feminine, but not both. However, Bem (1975) argued that individuals may instead be androgynous, that is both masculine and feminine, depending on the situation. Gender roles have been linked to psychological well-being, but studied almost exclusively among White Americans (Grimmell and Stern 1992). But one study by Napholz (1994) found that among 126 working Latina women from a Midwestern urban county, a majority of the women could have been considered androgynous. In addition, the research findings are mixed on the relationship between androgyny and life satisfaction. For example, Latinas who scored high on the masculine subscale were found to have higher scores on life satisfaction (Napholz 1994).

1.7 Social Support

The current study focuses on the most significant component of familism for Latinos, perceived social support from family (Sabogal et al. 1987). This may be due to the fact that Latinos emphasize mutual help and interdependence among family and the in-group (Sabogal et al. 1987). Also, disadvantaged minority groups, including Latinos, develop strong support networks with others in their ethnic group, including extended family and friends, as a way of coping with poverty and discrimination they experience (House et al. 1988).

Perceived social support has been argued to play a significant role in buffering and protecting against the negative impact of stressful events (Cohen and Willis 1985). Also, Meehan et al. (1993) contended that perceived social support might enhance adjustment by promoting positive affect, self-confidence, or a sense of personal satisfaction. Rodriguez et al. (2007) examined the relationships between several dimensions of familism, including family support, importance of family, and family conflict and the association of these to psychological well-being among 248 adults of Mexican origin. Placing high importance on family and perceiving high family support were positively associated with well-being (Rodriguez et al. 2007). Castillo et al. (2004) also found that among 247 Mexican American female college students in the West and South West United States, those who perceived high family support also perceived lower distress. Similarly, Solberg and Villareal (1997) surveyed 311 Latino students in a large university in the West United States and found that perceived social support was related to minimizing the negative effects of stress on reported physical and psychological distress. However, for students who reported low stress, perceived availability of social support was related to higher distress ratings. The authors speculate that individuals may feel obligated to provide support for others when the individuals, themselves, are not in need of support. Although these studies provide some insight to the benefits of perceived social support on the well-being of Latinos, further research in this area is clearly needed.

1.8 Summary and Limitations of Previous Literature

Mexican and Mexican American women experience multiple minority statuses, such as being an ethnic minority, being poor, and being female. Although Mexican and Mexican American women lack power and privilege that places them at risk for higher levels of stress and mental illnesses, this population also possesses strengths, cultural values, and experiences that are often overlooked. Consistent with Multiracial Feminist Theory and SWB research, researchers must focus on the strengths and positive attributes of this population in order to prevent further stigmatization and additional incidents of mental illness.

In reviewing the literature, several limitations were seen throughout the studies. First, acculturation was often measured with proxy measures, such as number of years living in the U.S. and language spoken at home (Gonzalez et al. 2007; Guinn and Vincent 2011; Lang et al. 1982; Leaper and Valin 1996), as opposed to using a validated measure. Secondly, the instruments used to measure gender roles were normed with White American college populations (Grimmell and Stern 1992). Because Latinos have more traditional gender roles than White American individuals (Baca Zinn 1976; Leaper and Valin 1996; Raffaelli and Ontai 2004), these instruments may not accurately measure gender roles among Latinos. A third limitation was that studies rarely focused on individual subgroups among Latinos (Raffaelli and Ontai 2004; Umaña-Taylor et al. 2009; Vera et al. 2011), which indicate an underlying assumption that all Latino subgroups are the same. Since very few psychological studies have examined SWB among low socio-economic status Latina women, and given the multiple minority statuses that Latinas experience, the target population for the present study is Mexican and Mexican American women of low SES.

1.9 Research Questions and Hypotheses

Based on the research of well-being, subjective well-being was operationalized as two separate constructs in this study: life satisfaction and positive affect (Diener 1984). Therefore, the present study addressed the following research questions: (1) What are the predictors of life satisfaction among low socioeconomic status Mexican and Mexican American women? (2) What are the predictors of positive affect among low socioeconomic status Mexican and Mexican American women?

Based on the literature, the following hypotheses were tested:

  1. 1.

    Identifying with both Mexican culture and the dominant culture (bicultural), greater ethnic identity, stronger androgynous gender role identity, and greater perceived social support from family will predict greater life satisfaction among Mexican and Mexican American women with low socioeconomic status.

  2. 2.

    Among the predictors, social support from family will be the most significant predictor of life satisfaction.

  3. 3.

    Identifying with both Mexican culture and the dominant culture (bicultural), greater ethnic identity, stronger androgynous gender roles, and greater perceived social support from family will predict more positive affect among Mexican and Mexican American women with low socioeconomic status.

  4. 4.

    Among the predictors, social support from family will be the most significant predictor of positive affect.

2 Methods

2.1 Participants

The 194 participants in the study were females living in California, who self-identified as either Mexican or Mexican American, were 18 years or older, were proficient in English, and identified as low socioeconomic status. Participants’ age ranged between 18 and 74 years (M age = 30, SD = 10.5). A majority of the participants had completed a high school education (48 %). Forty-one percent of participants reported having higher than a 10th grade education but did not specify their highest grade completed. Only one participant reported having a college degree. Twenty-eight female participants (14 %) were born in Mexico and 166 female participants (86 %) were born in the U.S. Of those born in Mexico, the mean age at time of immigrating to the U.S. was 9 years old. Only 3 females (2 %) reported living in Northern California and 191 (98 %) reported living in Southern California. One hundred twenty-three (63 %) female participants reported being single, 46 participants (24 %) reported being married, and 26 participants (13 %) reported cohabitating. One hundred eight (56 %) female participants reported they did not have children and the average household size was four.

To determine low socio-economic status (SES), the California state income limits for the Section 8 Housing Voucher Program (Bates 2013) was used. Participants were included in the study if they met the criteria for California’s definition of low SES, which is having 80 % or less of the Median Family Income (MFI) estimates for Los Angeles County (Bates 2013). The majority of the participants (66 %) resided in Los Angeles County. In order to determine English proficiency, participants needed to indicate on the demographic questionnaire that they were proficient in the English language and that they had completed the 10th grade.

2.2 Measures

2.2.1 Demographic Questionnaire

The demographic questionnaire created by the researchers asked participants to provide information regarding their age, ethnicity, income, size of household, generational status, marital status, parental status, highest level of education completed and English language proficiency.

2.2.2 Short Acculturation Scale for Hispanics

Acculturation level was measured with the Short Acculturation Scale for Hispanics (SASH; Marin et al. 1987). The SASH-Shortened Version consists of a 4-item, self-report measure scored on a 5-point Likert scale, ranging from 1 (Only Spanish), which indicates low level of acculturation, and 5 (Only English), which indicates a high level of acculturation. Sample items on the SASH include, “In general, what language(s) do you read and speak?” and “In what language(s) do you usually think?” Scores are converted into mean scores (Marin et al. 1987). The SASH was validated with Mexican Americans and Central Americans and showed similar results for both groups (Marin et al. 1987). The alpha coefficient for scores on the scale was .92, indicating adequate reliability (Marin et al. 1987). In the current study, scores on the SASH-Shortened Version had strong internal consistency (Cronbach’s alpha = .86).

2.2.3 Multi-Group Ethnic Identity Measure: Revised

Ethnic identity was measured with the Multi-Group Ethnic Identity Measure—Revised (MEIM-R; Phinney and Ong 2007). The MEIM-R is a revised version of the MEIM that assesses ethnic identity across diverse ethnic groups (Phinney and Ong 2007). The MEIM-R consists of a 6-item self-rating scale scored on a 5-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). A higher average score indicates a higher level of ethnic identity with the respondents’ identified ethnic group while a lower average indicates a lower level of ethnic identity with the respondents’ identified ethnic group (Phinney and Ong 2007). An example of an item on the MEIM-R is “I have a strong sense of belonging to my own ethnic group.” Phinney and Ong found strong scale score reliability for the MEIM-R (Cronbach’s alpha = .81) and Yoon (2011) found acceptable reliability when using an ethnically diverse college population in California (n = 289; 34.3 % Latino). Cronbach’s alpha in the current study was .88 for scores on the combined MEIM-R scale.

2.2.4 Bem Sex Role Inventory

Gender role was measured with the Bem Sex Role Inventory (BSRI) short form (Bem 1975). The BSRI short form (Bem 1981b) consists of 30 items, with 10 items on each of the three scales (feminine, masculine, and neutral). Participants rate each adjective on how it describes them using a 7-point Likert scale ranging from 1 (never or almost never true) to 7 (always or almost always true). Sample characteristics (items) on the BSRI include “affectionate” and “independent.” A median-split scoring procedure is used for scoring and classification (Bem 1981a). A median score is calculated for the femininity and masculinity subscales (androgynous, masculine, feminine, and undifferentiated; Bem 1981a). Androgynous is the result of scoring above the median in both masculine and feminine categories. Masculine and feminine categories are the result of scoring above the median in one gender and below the median in the other. An undifferentiated score is the result of scoring below the median in both masculine and feminine categories (Bem 1981a).

The BSRI appears to be a valid discriminator between the sexes for different ethnic groups in the U.S., including Mexican Americans (Reed-Sanders et al. 1985; Zeff 1982). Additionally, scores on the BSRI short form were found to have good internal consistency with a Cronbach’s alpha of .84 and .85 for the Femininity scale and Masculinity scale, respectively (Bem 1981a). Bem (1981a) also found strong test–retest reliability for the BSRI short form for the Femininity scale (r = .85) and for the Masculinity scale (r = .91). In the present study, scores on the BSRI short form showed strong reliability for the Femininity scale (Cronbach’s alpha = .89) and the Masculinity scale (Cronbach’s alpha = .79). In the present study, the sample of 194 Latinas was grouped into undifferentiated gender identity (n = 62), masculine gender identity (n = 38), feminine gender identity (n = 38), and androgynous (n = 56), based on scores on the BSRI short form. The groups were coded as follows: 1 = undifferentiated; 2 = masculine; 3 = feminine; and, 4 = androgynous.

2.2.5 Multidimensional Scale of Perceived Social Support

Perceived social support was measured with the Multidimensional Scale of Perceived Social Support (Zimet et al. 1988). The MSPSS distinguishes perceived social support from three sources, including family, friends, and a significant other. In the present study, only the family support subscale was used for analyses. The MSPSS consists of 12 items rated on a 7-point Likert scale, ranging from 1 (strongly disagree) to 7 (strongly agree). The MSPSS produces a total and subscale scores ranging from 1 to 7, with higher scores indicating greater levels of perceived social support. A sample item on the MSPSS is “I get the emotional help and support I need from my family.” The MSPSS was validated using a college population (n = 275; Zimet et al. 1988). Zimet et al. (1988) reported strong internal consistency (Cronbach’s alpha = .87) and strong reliability (Cronbach’s alpha = .88) for scores on the Family subscale. The MSPSS has been used cross-culturally (Canty-Mitchell and Zimet 2000; Dahlem et al. 1991) and Dahlem et al. (1991) also found scores on the MSPSS to have good internal reliability, (Family Cronbach’s alpha = .90). In the present study, Cronbach’s alpha for scores on the MSPSS Family subscale was .90.

2.2.6 Satisfaction with Life Scale

Global life satisfaction was measured with the Satisfaction with Life Scale (SWLS; Diener et al. 1999). The SWLS measures an individual’s global assessment of his or her own life. The SWLS consists of five items rated on a 7-point Likert scale, ranging from 1 (strongly disagree) to 7 (strongly agree). A score is obtained by summing the scores of all five items. Scores can range from 5 (low satisfaction) to 35 (high satisfaction). An example of an item on the SWLS is “In most ways my life is close to my ideal.” The SWLS is a widely used measure that has been shown to be both reliable and valid (Diener et al. 1999). In the current study, scores on the SWLC also showed strong internal consistency (Cronbach’s alpha = .92).

2.2.7 Scale of Positive and Negative Experience: Positive

Positive affect was measured with the Scale of Positive and Negative Experience—Positive (SPANE-P; Diener et al. 2009). The SPANE is a 12-item self-rating scale scored on a 5-point Likert scale ranging from 1 (very rarely or never) to 5 (very often or always). Six items on the SPANE assess positive feelings and six items assess negative feelings. The SPANE can be divided into three separate scales that measure positive feelings (SPANE-P), negative feelings (SPANE-N), and overall affect balance (SPANE-B). An example of a positive feeling on the SPANE-P subscale is “happy.” An example of a negative feeling on the SPANE-N subscale is “sad”, and an example of an item that is overall affect balance on the SPANE-B subscale is “content.” For the purposes of this study, only the SPANE-P was used. The score for this scale can vary from 6 (lowest possible) to 30 (highest positive feeling possible). Diener et al. (2009) found strong scale reliability for scores on the SPANE-P (Cronbach’s alpha = .87). In the present study scores on the SPANE-P had a Cronbach’s alpha of .89.

2.3 Procedure

IRB approval to collect data for this project was granted by the researchers’ institution. Initial recruitment of participants using convenience and snowball sampling, including recruitment through email and social media (i.e., Facebook), yielded 80 completed on-line surveys. Flyers were posted on bulletin boards of three community college campuses in the Los Angeles area. In addition, the primary researcher utilized a research panel from Qualtrics (a survey management company) to recruit an additional 114 participants who met the criteria of being Mexican or Mexican American, female, living in California, 18 years of age or older, and having low socioeconomic status. The research panel is representative of census data and includes over one million panelists. The electronic survey took about 15 min to complete. Consent to participate was provided electronically and required by all participants prior to beginning the survey. The first screen presented a consent form, which included information about potential risks and benefits of the study and the voluntary nature of participation. Contact information to three community college student health centers and a mental health hotline was provided at the end of the study in case the participants experienced any emotional discomfort or distress that surfaced from their participation in the study. Participants recruited through the initial phase were eligible to enter a drawing to win one of three $50 VISA gift cards. Participants recruited through the Qualtrics panel were compensated with credits to their Qualtrics account, which panelists could redeem for gift certificates.

3 Results

3.1 Preliminary Analysis

An analysis was done in order to compare the two non-random sample groups (self-recruited and participants recruited by research panel) used in this study across the six variables. For all variables except gender role identity, independent-samples t tests were done to compare the two sampled groups. For the six variables of interest, there were statistically significant differences between the two groups for ethnic identity t(192) = 2.95, p < . 01), perceived social support from family t(190) = 2.90, p < .00, life satisfaction t(176) = 1.94, p < .05, and positive affect t(192) = 3.73, p < .01. For the categorical variable of gender roles, a Chi square test showed there was a significant difference between the two groups in the distribution, χ 2 (3, N = 194) = 15.93, p < .00). There was also a significant difference in the two groups for both masculinity (p < .01) and femininity (p < .00) scores. Even though all participants in this study met the research criteria, individuals from these two samples differed in other demographic ways, which may have affected the results. These limitations are further considered in the discussion section.

3.2 Main Analyses

Before testing hypotheses, the assumptions of linear relationships, multivariate normality, no multicollinearity, and homoscedasticity were tested. All predictor and both criterion variables (life satisfaction and positive affect) were normally distributed (skewness < ±2.0 and kurtosis < ±2.0). Multicollinearity was within acceptable limits, with variance inflation factors (VIF) values under 5.0, for both life satisfaction and positive affect scores, for all predictor variables. Also, inspection of the scatterplots for both criterion variables demonstrated homoscedasticity.

To test the first and second hypothesis, a standard multiple regression analysis was performed. Life satisfaction was the dependent (criterion) variable, acculturation, ethnic identity, gender roles, and perceived family social support were independent (predictor) variables, and the demographic variables of age of participant, sample type (1 = self-recruited, 2 = research panel), marital status (1 = single, 2 = married, 3 = cohabitating), place of birth (1 = U.S., 2 = Outside of U.S.), and ethnicity (1 = self-identified Mexican, 2 = Mexican American) were included as control variables. All of these nominal demographic variables were dummy coded. The reference groups for each dummy coded variable are presented in Table 1. Results showed that none of the demographic variables (age, sample type, marital status, place of birth, and ethnicity of participant) were significant predictors in the model., However, stronger Mexican ethnic identity [ß = .26, p < .001] and greater perceived family social support [ß = .29, p < .001] were significant predictors of life satisfaction. This partially supported the first hypothesis [R 2 = .25, Adjusted R 2 = .20, F (12, 181) = 5.01, p < .001, Cohen’s f 2 = .33], with the model accounting for 25 % of the variance in life satisfaction. In addition, as predicted in hypothesis two, perceived social support from family was the most significant predictor in the model of life satisfaction, t (181) = 5.45, p < .001.

Table 1 Regression data for predictors for life satisfaction (N = 194)
Full size table

To test the third and fourth hypothesis, a standard multiple regression analysis was performed. Positive affect was used as the dependent (criterion) variable, acculturation, ethnic identity, gender roles, and perceived social support from family were the independent (predictor) variables, and the demographic variables of age of participant, type of sample (1 = self-recruited, 2 = research panel), marital status (1 = single, 2 = married, 3 = cohabitating), place of birth (1 = U.S., 2 = other), and ethnicity (1 = self-identified Mexican, 2 = Mexican American) were included as control variables. All of these nominal demographic variables were dummy coded. The reference groups for each dummy coded variable are presented in Table 2. Results showed that none of the demographic variables were significant predictors in the model. However, stronger Mexican ethnic identity (ß = .16, p = .02), greater perceived social support from family (ß = .37, p < .001), and androgynous gender roles (ß = .19, p = .01) were significant predictors of positive affect. The model of predictors accounted for 33 % of the variance in positive affect [R 2 = .33, Adjusted R 2 = .29, F (12, 181) = 7.53, p < .001, Cohen’s f 2 = .49], which partially supported hypothesis three. To further explore the influence of gender role identity on positive affect, a univariate analysis of variance was conducted. The participants were categorized into undifferentiated (n = 62), masculine (n = 38), feminine (n = 38), and androgynous groups (n = 56), based on their scores on the BSRI short form. Significant differences were found among groups for positive affect, F (3, 190) = 8.30, p < .001, η2 = .12. Participants who were classified as having a feminine gender role had significantly higher positive affect scores (M = 23.21, SD = 5.04) than participants with an undifferentiated gender role (M = 20.82, SD = 3.69; Mean difference = 2.39, p = .01) and participants with a masculine gender role (M = 20.76, SD = 4.44; Mean difference = 2.45, p = .01). Participants who were classified as having an androgynous gender role had significantly higher positive affect scores (M = 24.23, SD = 4.37) than participants with an undifferentiated gender role (Mean difference = 3.41, p < .001) and participants with a masculine gender role (Mean difference = 3.47, p < .001). Also, as predicted in hypothesis four, perceived social support from family was the most significant predictor of positive affect, t (181) = 5.45, p < .001.

Table 2 Regression data for predictors for positive affect (N = 194)
Full size table

4 Discussion

The purpose of this study was to examine whether acculturative status, ethnic identity, gender roles, and perceived social support from family were related to the subjective well-being of Mexican and Mexican American women of low SES. Of the four predictors of life satisfaction, only ethnic identity and perceived family social support were significant. As expected, perceived social support from family was the most significant predictor of life satisfaction. These results are consistent with previous research (Castillo et al. 2004; Cohen and Willis 1985; Rodriguez et al. 2007), which has shown that placing high importance on family and perceived high family support are positively associated with well-being. Our results suggest that Mexican and Mexican American women of low SES, who have a strong ethnic identity, and who perceive that they have a supportive family, are more likely to feel satisfied about their life in general.

Of the four predictors of positive affect, ethnic identity, gender roles, and perceived social support from family explained a significant amount of variance, which somewhat supports findings from previous research regarding the role of ethnic identity on positive affect (Vera et al. 2011). Our results also showed that individuals identifying with androgynous gender roles had greater positive affect scores than individuals identifying with undifferentiated and masculine gender roles, as expected. These findings make sense in that identifying with some kind of gender role (i.e., the opposite of being undifferentiated) is important to subjective well-being (Golding 1990). Additionally, the relationship between androgynous gender role identity and greater positive affect is supported in the literature (Bem 1975; Napholz 1994). However, the present study also found that individuals with feminine gender identities had greater positive affect, which does not support previous findings (Napholz 1994). Bem (1974) argued that women who mainly endorse feminine characteristics displayed the most serious behavioral deficits consistent with poor social adjustment. Espin (1997) also argued that for Latina women who retain more traditional gender roles, confusion and higher levels of stress may arise due to contradictions between cultural norms in Latin America and the U.S. Yet, from the perspective of the Latino cultural value of marianismo, being feminine would be an expected and more acceptable identity for Latinas (Gil and Vazquez 1996; Humphrey 1944). Our other finding that perceived social support from family, a core component of familismo, was the most significant predictor of positive affect supported our predictions, and is also supported by the literature (Almeida et al. 2009; Rodriguez et al. 2007; Sabogal et al. 1987).

4.1 Limitations

There are a few limitations in the present study worth noting. Participants were females who self-identified as having a Mexican heritage, lived in California, identified as low socioeconomic status, spoke English, and were Internet users. These criteria limit generalizability to the larger Mexican and Mexican American population. In addition, a majority of the women who participated in this study identified as Mexican American (86 %); however, Mexican and Mexican American women have likely had different experiences related to acculturation and ethnic identity. Additionally, two different non-random samples were used in this study. One sample was recruited through various methods (i.e., social media, email). The second sample was recruited through a research panel (Qualtrics). Not only did the two samples differ in mean scores across five of the six variables (ethnic identity, gender roles, perceived social support from family, life satisfaction, and positive affect), the two samples could have differed in other demographic ways (e.g., religious affiliation, self-esteem, personality variables) that were not measured; therefore, investigation into other factors that affect subjective well-being is warranted.

4.2 Implications and Future Directions

The large presence of Latinos in the U.S., particularly in states such as California, Arizona, and Texas (U.S. Department of Commerce, Bureau of the Census 2010), make it imperative for mental health professionals to become familiar with both the barriers these individuals encounter as well as the cultural strengths they possess. It is important that mental health professionals have a macro level understanding of how women and men throughout society experience different forms of privilege and subordination, depending on their race, class, and gender (Baca Zinn and Dill 1996). The results from the present study can provide mental health professionals with an understanding of some of the variables that contribute to the subjective well-being of Latina women. For example, there should be focused efforts by mental health professionals to include family members in the treatment of Latino clients. In addition, the results of this study can inform interventions aimed at increasing SWB in Latinas, including interventions from Multicultural Feminist Theory. Comas-Diaz (1991) discusses several Multiracial Feminist Theory principles that are particularly applicable to Latinas and can assist mental health professionals. One principle includes assisting Latinas in understanding that differences are not deficiencies and clinicians should avoid pathologizing these differences. A second principle includes acknowledging the relevance of women’s context because women’s experiences in a given culture are socially constructed (Comas-Diaz 1991). Empowerment, the third principle, helps women acknowledge and deal with negative feelings related to their ethnic minority status, and to successfully negotiate their cultural adjustment (Comas-Diaz 1991).

On a community level, efforts can be made to celebrate Latino culture and promote positive views of Latinos. Community health fairs can incorporate family friendly events, where professionals are bilingual Spanish speakers and events are targeted to individuals of all ages, including children. Consciousness-raising groups can discuss what individuals in the group deem most important to them and can help foster well-being in Latinas by providing them a support system where they can share their experiences and discuss ways to promote well-being in themselves and in their community.

Given that the variance accounted for by the models in the present study was minimal (≤35 %), the findings indicate that additional variables could be contributing to the subjective well-being of Mexican and Mexican American women. Other variables that are worth exploring in future research are religious and spiritual beliefs (Castillo 1995), perceived control (Meehan et al. 1993), and perceived support from friends, a significant other, and the larger community (Solberg and Villareal 1997). Future studies should also examine Latino males, given the limited research that exists on well-being for males. In addition, because people differ in their identity, interests, and experiences across the different stages of life, longitudinal studies could provide valuable information about the different variables that contribute to the subjective well-being of individuals across different ages in life. Lastly, exploring the role of social context may prove beneficial when examining the subjective well-being of Latinos. The ethnic diversity of a community, as well as the resources that are available, are important factors to consider when examining the well-being of Latinos (Umaña-Taylor et al. 2009). It is clear that Latina women’s identity and experiences are complex, but better efforts should be made so that Latina women are supported rather than oppressed by society.