formative third order construct

[darg]

Hello,
I have a formative-formative-formative 3rd order construct for which I would like to use the disjoint two step approach, following the recommendations of Becker 2012 and Becker 2022.
However, I am unsure how to start: how to obtain the required lower order construct scores if I cannot run the model in SmartPLs, because the higher order variables (red in graph) have no indicators (since it is a hierarchical model). I probably need to do it in two stages anyway, first the second-order construct and then the same procedure for the third order construct, but the problem remains on how to obtain the needed LOC scores.
I hope you can help me unravel the knot.
Thank you

Graphpure.png (44.95 KiB) Viewed 475 times

[jmbecker]

If you want to do the disjoint approach you need to relate the lower-order constructs to your antecedents and outcomes of the higher-order construct as is described in the papers that you mention.

[HeleneColon]

Thank you so much.

[darg]

Thank you for the hint, it helped to start the analysis process!