The journal NATURE reveals peer-reviewing scam

Post Reply New Topic

How much cheating goes on in the scientific world is grossly underestimated by the general public and media. It's sad when a journal as prestigious as Nature announces this..


http://www.nature.com/news/publishing-t ... am-1.16400

I'd believe that there is cheating in the scientific world. It is very competitive there.

Nothing sad about cheats being exposed. Quite refreshing actually that a prestigious publication like Nature cares enough to report on it.

Yes, that too.

All the scientists involved will lose all credibility and all their submitted articles will be reviewed again by all credible journals. Many of their articles will be likely be redacted by the journals. They're careers as scientists are likely over.

Thanks for that, it's good news.

Found an excellent website called "Retraction Watch" and it has extensive archives (access to archives on the RHS of the opening page):

http://retractionwatch.com/2014/11/25/p ... r-reviews/

Today I found a scientific article which confirmed a niggling suspicion I have had about the claims made for some fMRI studies, a misgiving that I have posted about in other threads when the media repeats claims like "neuroscientist proves the Aspergers in caused by bigger this or smaller that". The over-claiming upset me, the proof often just wasn't actually there, the links between the results, conclusions and claims seemed to be only a result of reductionist tunnel vision. It upset me that the claims were accepted so uncritically too - by media, by Autism Speaks, by other scientists, by WP members. However I attributed it to wishful thinking on the part of the authors.

Fortunately, someone more astute than me noticed that the correlations were often suspiciously high. This applied to papers published in respected journals like Science, Nature, and others.

http://www.pashler.com/Articles/Vul_eta ... npress.pdf

You are welcome. I'll continue to keep my eye out, this is an area of considerable interest for me.

I think the example in the article is unusual. Most prestigious journals would normally not allow self reviews. However I've known for a number of years of a "old boys club" of senior academics in a discipline who review each others papers...

The journals weren't complicit - they didn't realise they were being duped. Subsequent investigations found that this wasn't just a one off sort of thing.

It's not just fakery in peer review that has slipped past the radar of these prestigious journals. There's been a degree of editorial blindness to the fact of unusually high correlations occurring in some fMRI studies - that perhaps should have raised suspicion prior to acceptance for publication, but did not. Science, Nature and others have published apologies.

Essentially the researcher abused the software for a vulnerability. In other words, we can thank the innovation of automating the peer review process for this scandal.

This study, on the extent to which published papers used erroneous analysis (undetected by the publishers) is also a bit of an eyebrow raiser:

http://www.nature.com/neuro/journal/v14 ... .2886.html

and neuroimaging studies were particularly prone to this particular comparison error:

'As noted by others5, 6, the error of comparing significance levels is especially common in the neuroimaging literature, in which results are typically presented in color-coded statistical maps indicating the significance level of a particular contrast for each (visible) voxel. A visual comparison between maps for two groups might tempt the researcher to state, for example, that “the hippocampus was significantly activated in younger adults, but not in older adults”. However, the implied claim is that the hippocampus is activated more strongly in younger adults than in older adults, and such a claim requires a direct statistical comparison of the effects. Similarly, claims about differences in activation across brain regions must be supported by a significant interaction between brain region and the factor underlying the contrast of interest. For example, “Compared to non-moral dilemmas, the moral dilemmas activated only the insular cortex, suggesting that this area is uniquely involved in making moral judgments”. Identification of the significant response in the insular cortex does not imply that this region is uniquely or more strongly involved in making moral judgments than other regions. It merely implies that, although the null hypothesis has been rejected in this region, it has not been rejected elsewhere'.

I wish (!) the media could read this thread before touting flawed neuroimaging studies as big breakthroughs about us to the general public... often claims that in some way reinforce myths and stigma. I wish Autism Speaks was forced to copy the whole article out by hand 500 times as another penalty for their continual role as another tout of flawed claims!