@rockhopper72 Thanks for your response. I agree RCTs are not always good quality, and not always possible. Natural experiments involving twins are pretty close, but there aren't enough of them.
However, in the natural sciences you set out to test just one variable, and you should still aim to...