Hello:
I would be interested to use the results of a meta-analysis (correlations between 1st order constructs) to test 2nd order and structural model with PLS.
I think that EQS is not the proper procedure for the model that I want to test (although I would work with correlation matrices as input).
Is there any way to use correlation matrices as input method for PLS? (I know it is not CB-Sem ... but maybe there's a way to "simulate" the data from correlations).