Dealing with MIMIC models in SmartPLS

[mnabi]

Hi everyone

I have some constructs which have some formative and reflective indicators. Since we cannot evaluate MIMIC models directly in SmartPLS, I am looking for solutions. Any suggestion?
I also found these solutions from Petter, Straub, and Rai (2007):
"If a researcher is considering using a construct that seems to have some items that are reflective, while others are formative, there are approaches to address this issue a priori to data collection. First, the researcher could model the construct as having both formative and reflective items. This is actually
an ideal approach to ensure statistical identification of a structural model with formative constructs when using covariance-based SEM (we discuss this more later in this paper). Second, if the reflective items are measuring exactly the same facet of the construct, and the content validity of the construct would not be affected, all of the reflective items except one could be removed from the measure. The result would be a construct that has the same content validity (if the correlated items are interchangeable) as the original construct, and the researcher would be left with a purely formative construct. Finally, if the construct is important and a primary focus of the study, it may be most appropriate to model the construct as multidimensional. One could have separate dimensions for each aspect of the construct. This final option may increase the complexity of the model, but it may better describe the construct, measures, and underlying theory."
My question is: Is the second solution reasonable? Can I really use one reflective item beside other formative ones to have a purely formative construct? any idea?

Thanks

Petter, Stacie, Detmar Straub, and Arun Rai. 2007. “Specifying Formative Constructs in Information Systems Research.” Mis Quarterly 31 (4): 623–56.

[cringle]

Yes, the suggestion of using a formative construct and a reflective construct sound reasonable. But only use these two constructs with the same meaning in your model. If you add additional constructs, you extend the nomological net of your PLS path model and, thereby, change the content of the constructs.

In the SmartPLS sample project, you find an example of such two-level construct for the redundancy analysis.

Furhter references:

Chin, W. W. (2010). How to Write Up and Report PLS Analyses. In V. Esposito Vinzi, W. W. Chin, J. Henseler, & H. Wang (Eds.), Handbook of Partial Least Squares: Concepts, Methods and Applications (Springer Handbooks of Computational Statistics Series, vol. II) (pp. 655-690). Heidelberg, Dordrecht, London, New York: Springer.

Hair, J. F., Hult, G. T. M., Ringle, C. M., & Sarstedt, M. (2017). A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM) (2 ed.). Thousand Oaks, CA: Sage.

Best
Christian