actually Duck science is the only set of methods that do reliably separate out what works and what doesn't.

Your issue is with interindividual variability. The way to get around that is by varying conditions within the same person.