Jens Forster, yesterday, released a letter stating that he is innocent of scientific fraud, and of questionable research practices.
On the first, I don’t know one way or the other, but with the fraud spotlight on social priming at the moment, it would have been foolhardy of him to have published such research.
On QRP, I think he may have done, but on one QRP, he is not along.
As I published yesterday about whether journals encourage such QPR, I continue to believe that journals encourage researchers to only publish significant findings, and the best looking findings. How many of us have nonsignificant findings filed away that never see the light of day?
My suggestion to Dr. Forster would be to provide data files for these nonsignificant findings, which he admits to having had in the past. While such studies are unlikely to show the linearity problem present in his published work, things like the SDs could be compared.
Secondly, how many of us are guilty of recruiting participants until the difference is significant? While international labs might have the benefit of a legion of undergrad interns and unlimited supply of participants, meaning that they can do that inevitable power analysis that always says N = 120 no matter how many variables/conditions you are looking at, here in the UK, we tend not to have such luxuries. For many of us, our only option is to run the experiment until it is significant. It might be a QRP, but it is also a pragmatic one to funding shortages.
So, I think there are bigger picture problems here rather than just one rogue scientist. I think it is at least in Dr. Forster’s favour that independent labs have replicated his findings, something which has been the reason for uncovering fraud in other social priming experiments. If he has manufactured data, he was bloody lucky to continually hit home runs.