Probleme der Validierung mit Strukturgleichungsmodellen

This article evaluates the performance of structural equation models in validating measurement models for hypothetical constructs and deals with specific issues following from the way this methodology is typically applied in scale construction. In particular, controlling for various types of method effects is proposed as an alternative to eliminating indicators based solely on statistical criteria (e.g., internal consistency). -- Methodeneffekte ; reflektive Indikatoren ; Reliabilität ; Strukturgleichungsmodelle ; Validität