I guess a sample size of 3 is better than a sample size of 1, but how did this paper even get past peer review?