I read that pls-sem does not have a universal GoF, although I came across a formula that was later criticized by other authors. So is there any alternative measure I can use instead, or is it really necessary to report any GoF measure for pls-sem? Thanks.
Hey,
yes, there are approximative measures, like SRMR (Hair et al. 2014) and bootstrap-based exact measures of goodness of fit (Dijkstra & Henseler 2015). Both are implemented in SmartPLS and ADANCO. In ADANCO the implementation was more successful. In SmartPLS, you can't calculate a goodness of fit for a model with a mixture of common factors and composites.
Best regards,
Florian
Hair, Joe F., et al. "Common beliefs and reality about partial least squares: Comments on Rönkkö and Evermann." Organizational Research Methods 17.2 (2014): 182-209.
Dijkstra, Theo K., and Jörg Henseler. "Consistent partial least squares path modeling." MIS quarterly= Management information systems quarterly 39.2 (2015): 297-316.
Florian Schuberth, thanks for your answer. Much appreciated.
Dear Colleagues,
Good Day,
This subject is not my field but I found some papers that are related to the question, hope you find them very useful....
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling