PDF | Handbook of Structural Equation Modeling – Download

Assuming that metric invariance is found to hold, the next step in model evaluation is to place invariance constraints on the measurement intercepts. The combination of metric invariance and invariance in the measurement intercepts is known as scalar invariance (Steenkamp & Baumgartner, 1998) or strong factorial invariance (Meredith, 1993). Metric invariance is generally required before scalar invariance is evaluated because differences in factor loadings imply that the regressions of the measured variables on the factor scores are not parallel across groups. If these regressions are varying, group differences in the measurement intercepts are also likely because two regression lines with different slopes (i.e., factor loadings) will ordinarily also have different intercepts