His arguments don't apply to himself

The question asked at the end, about whether Dr. Ioannidis’ study is subject to its own conclusion, is fairly easy to answer.

The paper is really only relevant to studies that perform statistical analysis on a dataset, compute a "statistical significance" (p) for the dataset, and then asserts a claim based on the statistical significance. Ioannidis’ paper doesn’t do that.

Strictly speaking, it’s not a "scientific" paper at all. The argument he makes is entirely mathematical. He’s talking about the application of Probability Theory.

Anyway, I’ve been trying to digest his argument (and since the essay is actually a couple of years old, the counter-arguments) before I jump in too far, here. But I thought I’d at least throw this out there to see if anyone else is thinking about it.