Developmental Considerations and Acculturation of Children: Measures and Issues

This article presents the secondary validation of the Brief Acculturation Rating Scale for Mexican Americans-II (Brief ARSMA-II) for use with children—carried out using two samples of Mexican-descent children (ages = 9-11) from two states (N = 295). The Brief ARSMA-II was originally normed on adolescents and adults but has been validated and used with children. Ethnic identity development perspectives suggest that the interpretation of scores derived from acculturation measures normed on adolescents and adults may not extend accurately to children. Convergent validity and differential discrimination between groups were examined using scores on the Brief ARSMA-II; scores on an acculturation measure designed for the present study, the Things About Me (TAM); and traditional proxy measures of acculturation. Results from this study do not support the use of the Brief ARSMA-II with children. The importance of considering contextual effects in the interpretation of scores of children’s acculturation experience is discussed. NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be accessed by following the link in the citation at the bottom of the page. Hispanic Journal of Behavioral Sciences, Vol. 31, No. 1 (February 2009): pg. 57-72. DOI. This article is © SAGE Publications and permission has been granted for this version to appear in e-Publications@Marquette. SAGE Publications does not grant permission for this article to be further copied/distributed or hosted elsewhere without the express permission from SAGE Publications. 2 Acculturation models have evolved from early melting-pot perspectives, to the examination of cultural changes that occur in one or both groups that come into contact (Redfield, Linton, & Herskovits, 1936), to more recent incorporations of the psychological perspectives involved (Berry, 1980; Padilla & Perez, 2003; Teske & Nelson, 1974; Tropp, Erkut, Coll, Alarcon, & Vazquez Garcia, 1999). Researchers interested in exploring the disadvantages faced by minority populations often use acculturation models to understand further the dynamics between dominant and minority cultures (Born, 1970; Padilla, 1980; Williams & Berry, 1991). One of the measures used often to measure acculturation, as well as develop similar instruments to measure acculturation, is the Acculturation Rating Scale for Mexican Americans (ARSMA; Cuéllar, Harris, & Jasso, 1980). Cuéllar, Arnold, and Maldonado (1995) later modified the ARSMA to measure Mexican and Anglo-cultural orientation separately, and to result in four modes of acculturation: traditional, low biculturals, high biculturals, and assimilated (Acculturation Rating Scale for Mexican Americans II [ARSMA-II]). The ARSMA and ARSMA-II instruments were normed on adolescent and adult samples and included items designed to assess language preferences, ethnic identity, cultural heritage, and ethnic interaction (Cuéllar et al., 1995). More recently, Cuéllar (2004) developed an abbreviated instrument based on the ARSMA-II (Brief Acculturation Rating Scale for Mexican Americans-II [Brief ARSMA-II]) that maintains the acculturation construct while providing the researcher with brevity (Bauman, 2005). Cuéllar found that the language-based and peer-based items in the Brief ARSMA-II served as proxy measures for the excluded ethnic identity and cultural heritage factors among young adults, making their inclusion redundant and time consuming. The application of these three measures in the examination of physiological health (e.g., Campos, Dunkel Schetter, Walsh, & Schenker, 2007), mental health (e.g., Gamst et al., 2002), and academic achievement (e.g., Hurtado-Ortiz & Gauvain, 2007) has contributed to the understanding of the dynamic processes that can often result in deleterious effects among adolescent and adult Latinos. Purpose of the Study To extend the understanding of the impact of acculturation, researchers have attempted to develop and validate acculturation measures for children (Bauman, 2005; Martinez, Norman, & Delaney, NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be accessed by following the link in the citation at the bottom of the page. Hispanic Journal of Behavioral Sciences, Vol. 31, No. 1 (February 2009): pg. 57-72. DOI. This article is © SAGE Publications and permission has been granted for this version to appear in e-Publications@Marquette. SAGE Publications does not grant permission for this article to be further copied/distributed or hosted elsewhere without the express permission from SAGE Publications. 3 1984). As useful as it may be to apply the acculturative trajectory to children, Phinney’s (1992) ethnic identity development perspectives suggested that the interpretation of scores derived from acculturation measures used with adolescents may not be applicable to children. To determine whether acculturation measures reflect the same construct among children as among adolescents and adults, I examined the evidence of convergent validity and differential discrimination between groups using scores on the Brief ARSMA-II and scores on an acculturation measure developed for the present study designed to assess a child’s cultural preferences, the Things About Me (TAM). I also examined whether the relationship between traditional proxy measures of acculturation and scores from each of the two acculturation measures used in the present study were robust across different samples of children of Mexican descent. Finally, I determined whether traditional proxy measures of acculturation are indeed accurate proxies of acculturation. Developmental Perspectives Acculturation and ethnic identity are constructs that have been treated as orthogonal in some studies and interchangeable in others. Perhaps most accurately, acculturation and ethnic identity development can be described as interrelated (Cuéllar, Nyberg, Maldonado, & Roberts, 1997) and occurring simultaneously during adolescence (Phinney, 1992; Spencer & Markstrom-Adams, 1990). Because ethnic identity is treated as an essential component of acculturation in the various versions of the ARSMA (Cuéllar et al., 1995), ethnic identity development considerations must be applied in the interpretation of scores derived from all three instruments. Like most acculturation measures, the ARSMA, ARSMA-II, and Brief ARSMA-II rely on the self-reporting of preferences regarding cultural behaviors and, as such, function on the assumption that respondents have a preference that has resulted from the internalization and discernment of influences. For children, the limitations of assessing level of acculturation include developmental factors (e.g., the trajectory of identity development) and the dependence on selfreported preferences that may not be autonomous. Namely, although children may explore precursors to ethnic identity during middle childhood, they do not develop ethnic identity until late adolescence (Aboud & Doyle, 1993; Phinney, 1992). Thus, one of the obstacles in NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be accessed by following the link in the citation at the bottom of the page. Hispanic Journal of Behavioral Sciences, Vol. 31, No. 1 (February 2009): pg. 57-72. DOI. This article is © SAGE Publications and permission has been granted for this version to appear in e-Publications@Marquette. SAGE Publications does not grant permission for this article to be further copied/distributed or hosted elsewhere without the express permission from SAGE Publications. 4 attempting to measure acculturation among children lies in the ways in which scores are interpreted and used (i.e., validity). Phinney (1989, 1992) delineated the ways in which ethnic identity development changes across the developmental spectrum from adolescence to adulthood. Attributing her theoretical framework to the work of Erikson (1968) and Marcia (1980), Phinney explained that ethnic identity is rooted in the examination and challenge of attitudes (i.e., a developmental crisis). It begins with a period in which children give ethnicity little, if any, conscious thought and progresses to an exploration of the ways in which their ethnic group differs from others. During the final stage of ethnic identity development, individuals who successfully resolve their preceding challenges come to terms with who they are in terms of ethnicity. Traditional Proxy Measures of Acculturation for Validation Some researchers use language and cultural behaviors to derive level of acculturation (Cuéllar et al., 1980; Cuéllar et al., 1995). Other researchers assert that level of acculturation is contingent on the amount of exposure to the dominant culture, and thus refer to generational status or place of birth as proxy measures of acculturation (Ryder, Alden, & Paulhus, 2000). Although measures of acculturation have traditionally used these and other proxy measures in validation procedures, some assert that this practice may be problematic. Measures that reflect behavioral components of acculturation (e.g., language) tend to exclude affective components that are an integral part of an individual’s acculturation process (Tropp et al., 1999). Moreover, Tropp et al. (1999) asserted that the amount of exposure to the dominant culture one has had is quite distinct from the sense of belonging one may have toward the dominant culture. It has been argued that proxy measures do not measure acculturation, but exposure to cultural behaviors, and that reliance on proxy measures can create validity issues (Matsudaira, 2006). Some, consequently, have recommended that acculturation measures move away from proxy measures given the limitations of relying on isolated dimensions that are only fragments of an individual’s acculturation experience (Cabassa, 2003; Matsudaira, 2006). When acculturation instruments developed for children rely on proxy measures that are imposed (e.g., language acquisition NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be accessed by following the link in the citation at the bottom of the page. Hispanic Journal of Behavioral Sciences, Vol. 31, No. 1 (February 2009): pg. 57-72. DOI. This article is © SAGE Publications and permission has been granted for this version to appear in e-Publications@Marquette. SAGE Publications does not grant permission for this article to be further copied/distributed or hosted elsewhere without the express permission from SAGE Publications. 5 methods), resulting scores may not portray accurately an individual’s level of acculturation. In the case of children who are English language learners (ELLs) receiving disparate methods of language acquisition (e.g., immersion or bilingual education), the use of acculturation measures may reflect the language used in instruction rather than the child’s affective preference. Hence, those who interpret scores from acculturation measures that are designed to assess acculturation should consider internal validity threats that may provide competing hypotheses for the resulting scores. An underlying assumption of the Brief ARSMA-II is that language provides an accurate proxy for acculturation. Language may not be a behavioral preference inherent among language minority children but one that is influenced by the language acquisition policies of their respective states. The TAM instrument was designed to measure personal cultural choices that more closely align with internalized preferences common among children, but those were absent from the Brief ARSMA-II. Although there are many different acculturation measures, there should be a concordance between scores on acculturation measures if they both indeed assess a child’s acculturation. In the first validation procedure, I explored the convergent validity of the Brief ARSMA-II with TAM among Mexicandescent children in middle childhood (i.e., ages 9-11). Convergent validity between the Brief ARSMA-II and TAM would provide evidence in favor of the Brief ARSMA-II for use as acculturation measure that portrays children’s developmentally appropriate preferences despite its focus on linguistic and social preferences. In addition, both instruments should discriminate acculturation levels to a corresponding degree and result in similar classification scores among participants on both instruments. Proxy measures have traditionally correlated highly with acculturation scores when used with adults and adolescents (e.g., Unger et al., 2002). Construct validation for measures of acculturation among children have used socioeconomic status (SES) and bilingualism (Martinez et al., 1984) and language and geographic proximity (Bauman, 2005). In the second validation procedure, I correlated traditional proxy measures (generational status, place of birth, length of time living in the United States) with resulting acculturation scores on the Brief ARSMA-II and TAM for each group to evaluate further the construct validity of the acculturation measures NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be accessed by following the link in the citation at the bottom of the page. Hispanic Journal of Behavioral Sciences, Vol. 31, No. 1 (February 2009): pg. 57-72. DOI. This article is © SAGE Publications and permission has been granted for this version to appear in e-Publications@Marquette. SAGE Publications does not grant permission for this article to be further copied/distributed or hosted elsewhere without the express permission from SAGE Publications. 6 when used with children. Acculturation measures should be robust in terms of generalizability within the populations for which the measures were designed. In the present study, the correspondence between the acculturation measures and proxy measures should be similar between the two samples. In addition, to explore the utility of proxy measures, I explored the contribution made by each proxy measure in the relationship between proxy measures and Brief ARSMA-II scores.