ABSTRACT
This chapter provides an overview of approaches to estimate and compare latent means across cultures. It also provides an easy and nontechnical introduction to latent mean difference testing, explains its presumptions, and provides an empirical example using cross-cultural survey data. Using a mean and covariance structure analysis (MACS) approach, the estimation of latent means is straightforward. The MACS approach reflects the important difference between observed means and latent means. The chapter presents the syntax codes for LISREL and Mplus and a short description of the procedure in Amos. It discusses an overview of the different ways to identify the mean structure, and discusses mean estimation in the second-order factor analytic and multiple indicators multiple causes (MIMIC) modeling framework and when considering formative measurement models. The chapter illustrates the theoretical issues with an empirical example using data from the European Social Survey and concludes with a discussion of latent mean comparisons in MIMIC, second-order and formative models.
