2 ms·
It's incredibly expensive to take a population, then randomize them to two different diets, then see that they actually follow those diets that you have assigne
by amanaplanacanal 8y ago
It's incredibly expensive to take a population, then randomize them to two different diets, then see that they actually follow those diets that you have assigned them to, for any length of time. You can't just observe people after the fact, because people don't remember what they actually ate, and even if they did you don't know what all the confounding variables even are. I'm not sure how this can even get any better than it is.