The ways in which cultural groups vary from one another has long been a matter of everyday observation. The construction of valid and reliable measures of these differences remains problematic. Comparisons of survey responses rest on assumptions about the functional equivalence of translated items and of the assumptions that respondents make about the meaning of such surveys. This chapter explores the different possible levels of measurement equivalence. Psychological variables are most frequently latent rather than directly observable. Philosophers of science have discussed how best to address the challenges one faces when working with latent variables. If we are to claim that latent variables such as individualism or collectivism can account for particular differences between groups, specific counterfactual theorising is required as to the limiting circumstances under which such effects will or will not occur. At a more practical level, we can note that differences are frequently found in the characteristic survey response styles of respondents from different cultural groups, but decisions as to whether or when to discount such variations rest on answers to the more basic philosophical questions raised in the earlier section of this chapter.