I was lucky to have a University class where we studied studies - constructing them, running them, and evaluating them. We went through many historically import

I was lucky to have a University class where we studied studies - constructing them, running them, and evaluating them. We went through many historically important studies to understand them better. Our conclusion? Lots of studies are very flawed and lead society in the wrong direction for decades - even if double-blind and statistically sound. When we tried to reformulate the studies so they would have been correct, the most common problem was that it wasn't possible to reformulate them at the time they were designed - what to measure wasn't known, or there was no way to measure it, or inputs or outputs were not known to exist or known to be measurable. Really, I wish everybody had taken that class. I hear all the "fact-based" exhortations by people who don't really understand that they are trading one religion for another (but, hey, I believe in taking the best judgment at the time, based on the best data - just understand it's limitations and take those into account.)