Response styles in cross-cultural surveys: An overview on estimation and adjustment methods and empirical applications

Advances in Comparative Survey Methods: Multinational, Multiregional, and Multicultural Contexts (3MC)
,
pp. 477-499
,
Hoboken, NJ
,
Wiley
,
2019,

Timothy P. Johnson, Beth-Ellen Pennell, Ineke A. L. Stoop, Britta Dorer,
Liu, Mingnan, Tuba Suzer-Gurtekin, Florian Keusch, Sunghee Lee
ISBN: 978-1-11-888498-0 (print); 978-1-11-888496-6 (online)
DOI: 
10.1002/9781118884997.ch22
Abstract

Measurement equivalence or invariance is the prerequisite for cross‐cultural survey research. One of the most commonly studied measurement errors in cross‐cultural survey research is response style. This chapter focuses on two types of response styles, namely, acquiescent response style (ARS) and extreme response style (ERS), and demonstrates several empirical models for measuring the impact of these two response styles in particular. It presents and compares four empirical methods for measuring and adjusting response styles: OLS regression, confirmatory factor analysis (CFA), latent class analysis (LCA), and multidimensional unfolding models (MUM). In addition to the key predictor, the chapter explains several variables that have been correlated with response style, including gender, age, education, household income, and data collection mode. It reviews analytical models that are convenient tools for measuring the impact of response styles, when researchers attempt to make comparison of the substantive latent constructs.