I am working on new research and I need to test my theory using synthetic data composed of approx. 8 variables (1 of them is supposed to be the dependent variable) but ONLY 2 of them have significantly higher mutual information values than the others.
Any idea on how to do this?