Scientific fraud kills people. Should it be illegal? https://www.vox.com/future-perfect/368350/scientific-research-fraud-crime-jail-time
That psychopathic end result is quite chilling--all those deaths. I wondered why beta-blockers used to be seen as innocent drugs. Not so anymore. The docs and research establishment protect their own, if not a true cover-up, it's sweeping it under multiple rugs until disintegrated into old dust.
Currently the institutions are in on it. If you report that one of their employees is an academic fraud, they do everything they can to sweep it under the rug. Don’t even get me started on retracting papers. Almost impossible.
Publishing false, misleading, or dubious claims should be illegal and punished. The potential of research to do harm by disseminating false information is as great as its potential do good by disseminating true information.
Of course it should. And PACE is a strong example of why. It has been thoroughly discredited, and exposed as fraudulent in court no less, but the Lancet refuses to retract it, the medical establishment refuses to acknowledge its myriad failings, and people are still being given GET based on its false results across the globe.
I'm quite astounded by this example because part of me is thinking why on earth such a huge development could be based on one study given they must know how flawed it all is - whether it is a mistake or fraud. But then of course we have enough examples. And when I read Crawley et al (2013) which basically said it was running the same thing at the same time as PACE - in fact was it actually basically part of the PACE thing? - and it basically confirmed it didn't get the same result, and definitely no improvement in the function measure (slight difference in the Chalder fatigue scale) then isn't there also a problem with people who decide not to read the results that come out too? I have no idea for example if there might have been other papers on the beta-blockers with different results. It does need a regulator. I agree it is complex and needs to be something that isn't going to just encourage more burying and tactics. Or scapegoating. And there is a real issue with non-retracting and not currently having something that has to be added to the very front of the abstract saying what kind of issue something carries - if it is being 'kept' for historical value (so people can find 'that paper that turned out to be a fraud but 20yrs of x ended up being built on). But worse the notes on the papers that were then done based on those and cited them. Institutions covers a lot of possibilities. If we think of eg a university as an academic's employer then there is the issue of academic freedom. Which adds an additional dimension vs a different kind of workplace. There just seem to be a lot of cracks in the system that is supposed to somehow have things come out in the wash. Peer review/journals, the idea that there will be alternative sides to each debate when you have some subjects where they've separated themselves off from the other areas and don't include them in their literature review or allow critique. Methodological norms somehow get skirted for some areas and not commented on, with new things becoming the norm. I'd guess funders were supposed to be one gatekeeper too. But personally I think the independent regulator type thing needs to not just be focused on the most severe. There is obviously an issue that even goes down to UG level stuff of focusing on short deadlines and pressure to 'prove something' when all it creates is people having to churn out at a certain rate and not being able to even if they want to explore the methodologically correct way of doing things when your competitors are quicker and cheaper because they just use online surveys or whatnot. It needs to go further and be analysing the methods used and start being pulled into the quality measures like RAE in the UK, but also funding awards and so on to tackle to non-null rewards and so on. And start making eg 'finding something definitively null' also be seen as 'quite a finding', particularly when data is well shared instead of the whole paper being lots of flounce. You'd have assumed that in many areas these would theoretically be the turf of professional bodies, because indeed they have a profession to protect and the status of what that means. But that's no good if it isn't at a 'super-level' focusing on methodology and is allowed to be at the subject level so you have some bottom-feeder subject that just have lower standards and are fine with that, and/or have fuzzy lines that mean those in the regulator can worry about their career being impacted by their actions if future jobs might come from the area they regulate/the hierarchy overlaps. I guess the thing is that things need to sometimes begin as theories - although in science normally that would/should in itself involve various experiments and data that at least themselves were solid being put into said paper and someone then discussing possibility a, b, c could be possibilities explaining this and how these could be probed/tested. Not just come up with a theory based on beliefs and write a paper that isn't about how on earth can we rigorously look to disprove this, and the tougher that is the more 'sure' it might be that it has some water-holding ability. But just because the entire solution is 'too hard' doesn't mean that there aren't things that can and should be done. It's particularly a problem because it isn't just influencing the current generation of things, the way research has to build on previous literatures.
