Testing the Measurement Invariance of Nativism. (31st January 2019)

• Record Type: Journal Article
• Authors: Zhao, Yikai
• Other Names: Lipsmeyer Christine S. guestEditor.; Whitten Guy D. guestEditor.; Young Clifford guestEditor.
• Is Part Of: Social science quarterly. Volume 100:Number 2(2019)
• Page Start: 419
• Page End: 429
• Publication Date: 2019-01-31
• Subjects: Political science -- Periodicals; Social sciences -- Periodicals; 300
• DOI: 10.1111/ssqu.12594
• Languages: English
• ISSNs: 0038-4941

Abstract

Abstract : Objective: Survey questions regarding immigration have been frequently used to compare how mass publics think about nativism, but it remains unclear whether the resulting nativism measures are comparable across countries. In this article, we empirically examine whether nativism achieves scalar invariance—an important measurement feature for valid instrumentation—across very different country contexts. Method: We examine measurement invariance of a nativism scale using survey data from 22 countries representing 16, 096 individual respondents. To test for invariance, we utilize multigroup confirmatory factor analysis (MG‐CFA). Results: While the nativism scale fails to achieve full scalar invariance, we are able to establish partial scalar invariance, indicating the nativism scale is valid across many national contexts. Though highly correlated, the means of the latent nativist variable also provide different rankings than the means of an additive scale. Conclusions: With a few limitations, nativist sentiments can be meaningfully measured across political contexts. Across 19 of the 22 countries, respondents answered survey items on nativism in ways that suggest their answers reflect the same underlying latent construct.