2nd Order - MGA & the repeated indicators approach

[JosipJovic]

Hello everyone,

I have a 2nd order type II (ref-for) model. To estimate and validate 2nd order constructs, I have followed the extended repeated indicators approach which is described in "Sarstedt et al. (2019) - How to specify, estimate, and validate higher-order constructs in PLS-SEM". Up to now, the methodology and the results were totally fine.
Now, I would like to use a multi group analysis (MGA) to test differences in the structural model (e.g. gender differences), which need the LVS. However, I am unsure about the assignment of the calculated LVS to the specific cases from my sample (e.g. to case 34 with gender female) as I have two data sets: 1) the original dataset and 2) LVS. I have not found any resource on this. How do I assign the calculated LVS to the specific cases from my sample? How are the LVS data ordered?

Best regards,
Josip Jovic