Originally Posted by eribrav

Actually the authors of this did not make it any kind of study,(never mind placebo controlled, double blind) snowy.

This is what's called a data mining expedition. It's typically a way that grad students or fellows write really bad, useless scientific papers.

There was no study done to test a hypothesis here. What the authors did was take old logs from prior studies, then look at the data and find differences in the groups. You can do this with old data and if you just keep comparing different sets of responses, eventually you will find two that are statistically different from each other. It's completely different from actually having a hypothesis, designing an experiment to test the hypothesis, and analyzing the data your experiment generates.

This is pure bunk and would never be published in a quality peer-reviewed journal.