I am conducting a study with simple mediation model with two continuous predictors X1 and X2, one continuous mediator M, and one outcome Y. Participants would rate both X1 and X2 at the same time, using the same measurement scale. Because X1, X2, and M are all formative constructs, so I plan to run PLS-SEM to test the structural relationships of my model. One of my research questions is whether the indirect effect of X1 is stronger or weaker than that of X2 on Y through M. Which command in R will let me conduct this comparative analysis?
I tried to run PLS-SEM in R using the plspm package following the instructions of Ravand and Baghaei (2016). In the output, I can see the results of significance testing for each path. However, the results do not test the significance of the indirect effects, nor do they allow comparison of indirect effects. I have tried the Help section but to no avail. I wonder if it is possible in R to construct the bootstrapped confidence intervals of the indirect effects and their difference. This should be easy using PROCESS, but PROCESS does not work with formative scales, so I am all out of ideas. Your help is greatly appreciated.