Cross-Ethnic Equivalence of Socialization Measures in European American and African American Youth
Krishnakumar, Ambika, Buehler, Cheryl, Barber, Brian K., Journal of Marriage and Family
We examined the cross-ethnic equivalency of socialization measures developed primarily with European American families. Four aspects of measurement equivalence were assessed: conceptual, operational, scalar, and functional. Evidence of between-and within-group measurement equivalency of socialization measures was derived from youth reports of 500 European American and 134 African American individuals ages 10 to 18, using confirmatory factor analyses and item response theory analyses. Findings indicate that most individual indicators of socialization, with the exception of paternal psychological control and parent-youth conflict, demonstrated cross-ethnic equivalence. The findings also suggest that lax control is better represented as a multidimensional construct (leniency and laxness).
Key Words: adolescence, cross-cultural, ethnicity, measurement equivalence, parenting, socialization.
Researchers have long recognized the need to approach issues of family processes with cultural sensitivity (Gjerde & Onishi, 2000). Until recently, investigations of family processes in ethnic groups other than European American middle-class families have been few (McAdoo, 1993). A starting point for valid cross-ethnic research is to examine whether measures validated primarily with European American middle-class families can be generalized to youth and families in other ethnic groups (van de Vijver & Leung, 1997).
Comparative research presents the theoretical challenge of assessing whether differences or similarities should be expected across ethnic groups. This task is fundamental to disciplines such as cross-cultural psychology and anthropology, where the distinction between emic and etic approaches to comparing human groups is central (Yau-Fai Ho, 1994). Understanding socialization behaviors from an emic perspective suggests that differences in behaviors across European American and African American families occur because each of these ethnic groups embodies characteristics (e.g., values, customs) unique to its specific group membership (Ocampo, Bernal, & Knight, 1993). For example, African American families place greater emphasis on collectivism and cooperation (Gaines, Buriel, Liu, & Rios, 1997), which, along with their experiences of racial discrimination and prejudice (Harrison, 1985), might produce relatively unique socialization experiences. Deducing hypotheses from an emic paradigm entails specifying ethnic group differences on means or slopes (Deater-Deckard, Dodge, Bates, & Pettit, 1996).
Understanding socialization practices from an etic perspective suggests that there are no substantial differences between European American and African American families on the constitutional elements of socialization. From this perspective, interpersonal dynamics between parents and children are fundamental to being human, and therefore are not highly determined by race, culture, or geographic location (Whiting & Edwards, 1988). In addition, over time, the intermingling of different ethnic groups might produce similar rather than unique socializing processes because of acculturation (Rowe, Vazsonyi, & Flannery, 1994). Deducing hypotheses from this paradigm entails specifying mean and slope similarities across ethnic groups (Deater-Deckard et al., 1996).
In addition to addressing the issue of cultural generality or specificity in socialization, research on cross-ethnic equivalency provides information regarding measurement that can be used to evaluate inconsistent findings across studies and to inform future research. Hui and Traindis (1985) outlined four types of equivalence: (a) conceptual or item equivalence (items have the same meaning and indicative behaviors across cultures), (b) operational equivalence (items are operationalized using the same method and response scale across cultures), (c) scalar equivalence (scores on items assessing a construct are similar in strength across cultures), and (d) functional equivalence (scores of a construct have similar correlates across cultures). …