ABSTRACT
When immigrant minorities engage in a new cultural context, their patterns of emotional experience come to change – a process we coined emotional acculturation. To date, research on emotional acculturation focused on the antecedents and consequences of changes in minorities’ fit with the new culture. Yet, most minorities also continue to engage in their heritage culture. Therefore, the current research investigated which personal and situational factors afford minorities to maintain emotional fit with their heritage culture. Two studies compared the emotional patterns of Korean Americans (n = 49) with those of Koreans in Korea (n = 80), and the emotional patterns of Turkish Belgians (n = 144) with those of Turks in Turkey (n = 250), respectively. As expected, we found that although minorities did not fit the heritage emotional patterns as well as participants in their home countries, spending time with heritage culture friends and interacting in heritage culture settings explained within-group differences in minorities’ heritage culture fit. Therefore, the current research shows that minorities’ emotional patterns are not only cultivated, but also activated by their interactions in different socio-cultural contexts. Moreover, it provides further evidence for cultural frame-switching in the domain of emotion.
Acknowledgements
We thank Berna Çoker and her colleagues from the Buça Eğitim Facultesi at the Doküz Eylül Üniversitesi in Izmir and Joke Van Eylen for their support during our data-collection in Turkey. We also thank Kimin Eom and Hyewon Choi for collecting the data in Korea. The data and SPSS-syntax that support the findings of these studies are available from the corresponding author, Jozefien De Leersnyder, upon reasonable request.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 Korean Americans are more highly educated and better employed than Turkish Belgians (FOD Werkgelegenheid, Citation2009; Terrazas, Citation2009) and face a racial rather than a religious “divide” whereas the opposite is true for Turkish Belgians (Alba, Citation2005; Yoo & Chung, Citation2009). Moreover, both groups navigate very different acculturation environments because immigration patterns, discourses, institutions and policies differ substantially between the United States and Belgium (Kosic & Phalet, Citation2006; Van Acker, Citation2012).
2 In both Study 1 and Study 2 we employed a Simultaneous Component Analysis (SCA; De Roover et al., Citation2012) to assess structural equivalence of the emotion data. This analysis provides insight into (i) whether one common factor solution can be used across the different samples under study and (ii) which items load on different factors, implying that they are not structurally equivalent and hence, differently understood across the cultural groups. For instance, in Study 2, “feeling resigned” loaded on the negative autonomy-promoting emotion component in the Belgian sample, but loaded on both the negative autonomy-promoting and the positive relatedness-promoting components in the Turkish Belgian samples. Personal conversations with Turkish Belgian participants revealed that “resigning” can be understood as embracing “kismet” (i.e., the Turkish concept of faith), which has a positive connotation in the Turkish cultural context. Although these cultural differences are interesting in itself, we removed items like this from the emotional patterns before calculating “fit”, because any (cultural) difference in intensity on these items may be due to their different meaning, while “fit” is about the (cultural) differences in patterns of intensity across emotions that have similar meanings across the groups under study.
3 Although one item did not load on the expected factor (Maintaining Values and Traditions), Cronbach’s alpha got worse (instead of improved) when removing this item from the scale. Therefore, and in keeping with both Study 1 and the 2011 paper, we retained this item.
4 Minorities’ attitudes toward social contact with Turks were highly correlated with their explicit statement of having Turkish friends (r = .435, p ≤ .001), but uncorrelated with the more implicit measure of speaking Turkish among friends (r = .061, p = .481); the latter two scales were moderately associated (r = .171, p = .047).