Discussion in 'Other Health News and Research' started by Dolphin, Dec 31, 2013.
Just on problems in scientific research in general. Links to a nice spoof.
Do read the paper this ultimately links to: http://www.dcscience.net/lawrence-current-biology-2007.pdf
Thanks, it's a good read, eg:
What has suprised me is the extent to which the top journals, in their bid to find eyecatching findings, publish poor studies too (alongside more great stuff than lesser jourals). At least one top scientist is taking a stand:
This shows that the problem is by no means confined to psychology or CFS research. The PubPeer link above is about a biomedical paper on the detection of disease biomarkers where the data presented imply the detection of single molecules of biomarkers. Even if this were the case, numerous posters point out there should be far more noise in the data: if you have multiple samples and an average of one biomarker molecule per sample, some will actually have 0, others 2 or 3 etc. Yet the data didn't show such noise, and the Nature editor contacted about this didn't seem to want to discuss the problem seriously. I know that sort of problem has come up in published CFS research too.
Perhaps the biggest difference is that in CFS it's left almost entirely to patients to challenge poor research (XMRV excepted), as biophile notes on another thread:
I know scienitists who say that in their fields strong criticism is routine from fellow-professionals. While this can descend into people gaming the system (eg reviewer's unreasonably blocking a competitor's paper) I think it's a much healthier state of affairs.
Thanks. This from the Peter Lawrence paper [pdf above] sums up the dilemma faced by many scientists forced to chase publication over discovery:
You can also try a Google Site Search
Separate names with a comma.