ABSTRACT
Cross-cultural analysis is all about comparing groups. Whether the groups being compared are countries, ethnicities, religions, or scientific fields, the goal is always to assess whether these groups differ in some respect and to interpret those differences in substantive terms. This chapter discusses a complementary approach to measurement invariance evaluation: sensitivity analysis using the "expected parameter change" in the parameter of interest, or expected parameter change (EPC)-interest. The EPC-interest in this context is a measure of the hypothetical change in a free parameter of substantive interest if a cross-group measurement invariance restriction were freed. The chapter focuses on sensitivity analysis as a method of dealing with the problem that measurement invariance is almost never exactly true. Sensitivity analysis addresses the questions directly by evaluating the impact of freeing cross-group equality restrictions on the parameters of substantive interest.
