We began the analysis by calculating and assessing
the reliabilities and validities of three country-specific
measurement models. Formative loadings of the six
primary dimensions on SRSS were in excess of .60, as
shown in Figure 1. For the reflective dimensions of
the high-order SRSS construct, all loadings were in
excess of .70. However, in the Chinese dataset, three
reflective measures cross-loaded, whereas this crossloading
did not occur for the American and Emirate
samples. This mixture of results could plausibly indicate
measurement invariance across the national contexts
of interest (i.e., that the SRSS construct itself
means theoretically different things in the Chinese
context than in the others), and so we further investigated
this possibility. In order for the adapted scales
to have validity for our current purposes, we first had
to establish conceptual invariance across the national
contexts of study.