Thanks for posting this,
@Eagles.
Actually, its not too bad as a defence of the PACE trial - given that its tough to make an argument here to save it. Some of the points are actually reasonable:
The first bit is true - we did use a tough method of correction. We could have used 3 comparisons.
However, the second bit of that sentence is blindingly stupid - 52 weeks is the primary endpoint of the trial. It's the
primary endpoint. It's the PRIMARY endpoint.
We could have presented odds ratios (obviously not confidence intervals, they're only for continuous data). But this isn't what you said you'd do in the protocol. We did
what you originally said you'd do. That was the whole point.
Well, I don't think that's defensible. We carefully examined every published statement made in justification of the changes. No more can be expected, especially not from such a hostile group. Look what it took to get that little data sample we worked on - imagine trying to get them to actually answer our questions!
Translation: even if our study wasn't that impressive, there's all those great other positive studies out there.
Are you referring to the previous, lower-quality poorly controlled studies that gave inflated effects due to their lack of appropriate controls - the ones you were trying to supersede with your better controlled "definitive" study? Its a bit disappointing to be back there in 2007 again. Alas that's the way of artefactual effects in Psychology. The more you control against them, the more they disappear!
For me, I love a good academic argument. Its the stuff that moves science forward. This defence is not a bad effort, but pretty easy to counter. I wish I had a cleverer adversary. That would be much more fun.