Abstract
This study examined the structure of the Spanish version of the Multidimensional Scale of Perceived Social Support (MSPSS) among a sample of 122 undocumented Hispanic immigrants. The component structure of the scale was examined using principal components analysis (PCA). Contrary to the findings of the original three-subscale structure (Friends, Family, and Significant Other), the results yielded a two-subscale structure (Friends and Family/Significant Other), indicating a merging between two of the subscales. Internal consistency was examined and found the Spanish version of the MSPSS reliable. It appears that this version of the MSPSS is useful for undocumented Hispanic immigrants, given that the measure be used with two subscales as opposed to three.
Hispanics face many challenges while migrating to the United States including language barriers, possible discrimination, and pressures to acculturate. Undocumented Hispanics may face even more problems due to their unique legal status. Specifically, this group is often prohibited from obtaining documentation necessary for adequate functioning, such as a valid driver’s license, social security card, health care, and employment. Thus, undocumented Hispanics must find ways to cope with pressures associated with the acculturation process.
One resource that Hispanic immigrants use to cope with these issues is social support. Social support can come from friends, family, or peers and is an important factor in their overall psychological well-being. Latinos have been known to emphasize the importance of close family relationships, also known as familialism (Freeberg & Stein, 1996; Marin, 1993). Such familial social support may serve as a protector against psychological maladjustment among Hispanics. Research further suggested that for Latinos, possessing warm, interpersonal relationships is an important cultural value and source of social support (Campos et al., 2008).
Research notes that caution should be taken while choosing measures for the Latino population, and researchers should ensure that the measures possess adequate psychometric properties for this group (Edwards, 2004). Zimet, Dahlem, Zimet, and Farley (1988) originally developed the Multidimensional Scale of Perceived Social Support (MSPSS) to assess perceived social support among American undergraduate college students. In their study, a three-subscale factor structure emerged: Family, Friends, and Significant Other. The scale has been validated on a number of populations including Mexican adolescents (Edwards, 2004), Chinese adolescents (Cheng & Chan, 2004), Thai populations (Wongpakaran, Wongpakaran, & Ruktrakul, 2011), African American adolescents (Canty-Mitchell & Zimet, 2000), as well as Other populations.
While the validity of the MSPSS has been demonstrated for a variety of populations, including Mexican adolescents, no study has investigated its utility among undocumented Hispanic immigrants in the United States. MSPSS has been validated on Spanish-speaking populations in Spain (Landeta & Calvete, 2002), but not on Spanish-speaking populations within the United States. The present study examines the factor structure and reliability of the Spanish version of the MSPSS for undocumented Hispanic immigrants currently residing in the United States.
Method
Participants
Participants consisted of 122 undocumented Hispanic adults (70 males and 52 females), ranging in ages from 18 to 58 with an average age of 33.66 years (SD = 8.24). Participants’ amount of time residing in the United States ranged from 5 months to 42 years with an average of 14.99 years (SD = 8.47). Furthermore, of these 122 participants, 18% reported living with friends (n = 22) and 82% reported living with family (n = 100).
Materials
MSPSS
The MSPSS is a 12-item self-report measure that analyzes an individual’s perception of available social support (Zimet et al., 1988). The scale consists of three subscales that measure different sources of support: Family, Friends, and Significant Other. Each subscale is rated from 1 (very strongly disagree) to 7 (very strongly agree) and included questions such as “There is a special person around when I am in need” and “My family really tries to help me.” Each subscale has been found internally reliable with Cronbach’s alphas of .87, .85, and .91, respectively; furthermore, reliability of the total scale was .88 (Zimet, Powell, Farley, Werkman, & Berkoff, 1990). The present study used the Spanish version of the MSPSS considering the limited English proficiency of this population, which was translated into Spanish and then back-translated by two bilingual psychologists, originally yielding a Cronbach’s alpha of .89 on a Spanish population (Landeta & Calvete, 2002).
Procedures
The data for this study were collected as part of a larger study examining culture-specific correlates and their impact on depression among undocumented Hispanic immigrants. Original data were collected by making contact in Hispanic communities in various regions of Arkansas and Houston, Texas. We traveled to various businesses and churches in these communities where permission was solicited from authority figures (managers, pastors, supervisors, etc.) to inquire about possible interest in our research among their employees or congregation members. Interested participants were provided with informed consent regarding the study, and willing volunteers filled out a packet of scales, one of which was the MSPSS. Due to the sensitive nature of this population’s undocumented legal status, we ensured that all information was confidential and would not be shared with any outside sources in the informed consent. Furthermore, no names were requested from participants to ensure that results would not be connected back to them in any way. Participants were not paid for participation in the study.
Results
We first calculated the internal consistency coefficients of MSPSS based on the original three-factor model. The Cronbach’s alpha value for the full scale was .92. The three subscales yielded Cronbach’s alpha coefficients of .88 (Family), .91 (Friends), and .88 (Significant Other). These coefficients suggested that the Spanish version of MSPSS is a reliable measure of social support for undocumented Hispanic immigrants. However, this does not necessarily warrant its validity. To establish MSPSS’s validity for use with this particular population, we examined the structure of the Spanish MSPSS among undocumented Hispanic immigrants through exploratory principal components analysis (PCA). 1
Before performing PCA, we checked the appropriateness for conducting this analysis. The correlation matrix (see Table 1) was examined, and many of the correlations were above .30. Also, the Kaiser-Meyer-Olkin value was .82, which exceeded the recommended value of .6 (Kaiser, 1970, 1974). Furthermore, the Bartlett’s Test of Sphericity (Bartlett, 1954) was statistically significant, supporting the factorability of the correlation matrix.
Inter-Item Correlations for MSPSS (Spanish Version).
Note. MSPSS = Multidimensional Scale of Perceived Social Support.
PCA revealed two components with eigenvalues greater than 1, explaining 55.1% and 16.5% of the variance, respectively. Examining the scree plot, there was an obvious break after the second component. Utilizing Catell’s (1966), we decided to retain the two components for further examination. The maintenance of the two components was supported by the results of parallel analysis, which revealed that only the eigenvalues of two components were greater than those obtained from a randomly generated matrix of the same size (12 variables × 122 participants).
The two-component solution explained a total of 71.54% of the variance. Component 1 contributed 55.07% and Component 2 contributed 16.46%. To better interpret these two components, and because variables were correlated, we performed both a direct oblimin and varimax rotation, which revealed a number of strong loadings for both components. There were little to no differences for eigenvalues between these rotations. Therefore, we used the varimax rotation for clearer interpretation. The interpretation of the two components was not consistent with the three-factor structure found in the original development of the scale (Zimet et al., 1988). Instead, the family and significant other variables loaded on the first component and the friends variables loaded on the second component (see Table 2). There was a weak to moderate correlation between the two extracted factors (r = .46) accounting for only 16% variance. The results of the analysis do not support the original three-factor use of the Spanish MSPSS for undocumented Hispanic immigrants.
Pattern/Structure Coefficients Matrix for PCA With Varimax Rotation of Two-Factor Solution of Multidimensional Scale of Perceived Social Support Items (Spanish Version With English Items in Parenthesis).
Note. Major factor loadings are in bold greater than .50. PCA = principal components analysis.
To establish reliability for the two-component subscales, Cronbach’s alpha was computed with the new combined subscale (Family/Significant Other). Internal consistency was high yielding a coefficient of .93, revealing the two-subscale structure of the Spanish MSPSS as a reliable measure.
Discussion
We examined the structure of the Spanish version of the MSPSS using PCA among undocumented Hispanic immigrants. We found that the original three-factor structure (family, friends, and significant other) was not supported for our sample. Instead, a two-component model emerged with “friends” as one component and the combination of “family” (FAM) and “significant other” (SO) as the other. The loading of SO and FAM as one component is not surprising. Familialism, as mentioned earlier, is an important cultural value for many Latinos (Freeberg & Stein, 1996; Marin, 1993). Specifically, Latino families often consist of close-knit relationships among members and are not limited to immediate relatives. For Latinos, a family may be blended, extended, nuclear, or traditional. As such, the idea of a significant other may present some ambiguity while trying to distinguish between “significant” people in one’s life (SO) and family. Indeed, Zimet et al. (1990) noted the importance of exploring exactly what is meant by significant other for the particular population being studied.
There are both theoretical and practical implications of the present findings. The MSPSS may be a more proximate measure of social support if utilized with two subscales (FAM/SO and Friends [FRI]), as opposed to the original three found in the original study (FAM, FRI, and SO). Combining the two subscales produced a highly reliable coefficient (.93), which makes this scale utilizable for measuring social support with two subscales instead of one. Theoretically, it appears that sources of social support were more easily identifiable with family and friends, and not a significant other.
The present study possesses a few limitations. Our study did not distinguish between countries of origin among participants. Future research may replicate this study on Hispanic Americans to examine possible differences in factor loadings. Overall, the Spanish version of the MSPSS seems to be psychometrically reliable and valid for this particular population, given that the scale is utilized with two subscales rather than one.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.