Christian Welzel opened 9th LCSR International Workshop with talk on MGCFA
Within the framework of the 9th LCSR International Workshop “Comparative Social Research” with an honorary report on the topic “Measurement Equivalence? A Tale of False Obsessions and a Cure” was made by Christian Welzel (LCSR HSE, Leufana University Luneburg).
In the course of the study, the possibilities of the method of multigroup confirmatory factor analysis (MGCFA) are considered. The main idea of the MGCFA approach is that, since within a particular country, the variability of factor decisions in multi-complex constructs is noted, the national averages are not fundamentally comparable. As a result, these constructs are considered useless for intercultural research. The report proves that the interdependence of national averages and factor decisions is misunderstood, which in turn distorts the very logic of intercultural comparisons. In the case of the distribution of average values across countries over a wide range of the scale of the construct, which is inevitable in the case of cross-cultural studies, the variability of factor decisions is logical and typical due to the fact that factor decisions are directly dependent on the average values for countries. Therefore, the assumption existing in the MGCFA approach that country averages are incomparable due to the variability of factor decisions needs to be fundamentally revised in order to correctly understand the logic of cross-cultural comparison.
Thus, this study reinforces the conclusions of C. Welzel and R. Inglehart that the comparability of multi-complex constructs should be determined depending on their relationships with external variables, and not on the basis of the lack of variability of internal factors.