ABSTRACT
It is not only the standard model evaluation procedure that we suggest to change. e standard model (Meredith, 1993) that is used to assess the cross-national equivalence is awed too. e reason that the standard model is awed is because the unique factors in the common factor model are confounded with random measurement error and also that the common (substantive) factor is confounded with systematic measurement error. erefore the invariance restrictions will lead to wrong conclusions if both random and systematic measurement error components are not the same across countries. We suggest making a distinction between the unique components in the indicators and the random errors in the indicators, as well as between the systematic error component in the indicators and the common factor. e same was also suggested by Saris and Gallhofer (2007) and by Millsap and Meredith (2007), the latter two authors did, however, not elaborate on this, nor did they empirically make this distinction.
