When I first started evaluating research, I was able to deconstruct studies and show why they were bad left and right. I thought I was so smart, but then when

When I first started evaluating research, I was able to deconstruct studies and show why they were bad left and right. I thought I was so smart, but then when I started trying to do my own research it became obvious that we operate under the same limitations as everyone else -- how do we design the best imperfect study given the limitations of money, time, people, and geography? In that context, I think this study is great. It is always true that the results of a study in a narrow group do not extrapolate to a larger group and it is good to be aware. However, that doesn't mean it can't generate hypotheses for a larger group (which can then be studied). The researchers likely would not have had the pull to generate an equivalent amount of data and follow-up data for a world-wide population, but I'm glad they did the best they could rather than spend their day posting criticisms on HackerNews.