Hacker News Reader

1200 person study is a pretty big study. That said, the authors of the study themselves listed some pretty big caveats.

Then the problem is not with the creators of the trials, but those who are reporting them with insufficient characterizing information.

Those are some pretty big caveats...