I have 21 independent variables (IV) and these variables are causing the reliability construct (RC) and the reliability construct is affecting the performance of the process in terms of time and cost indices.

I used PLS-SEM to calculate the R2 (COEFFICIENT OF DETERMINATION). I got the value of 0.536. I had three level construct. First level IV's, second level construct Reliability, and third level time and cost indices.

In the next stage I regrouped the 21 (I Vs) into sub-groups with 2, 3 or 4 variables to each sub-group and I have 4 levels of the hierarchy. at first level IV in group, second level regrouped construct, third level Reliability construct, and at the fourth level time and cost indices.

Now my R2 is 0.496.

Kindly explain and if some references.