7 ms·

Springer retracting 107 papers published in Tumor Biology with fake peer reviews

98 points by mathgenius 9y ago

DanCarvajal 9y ago
Trust Science.
jMyles 9y ago
Wow, that's bonkers. I wonder if any of these papers has definitely convinced a decision-maker to pursue or avoid a particular course of treatment. IE, has someone literally died because of this? It seems likely, but these papers are only a year or two old. In any case, this kind of happening is why it's difficult anymore to see "peer review" as being tantamount to a verification that a conclusion was reached via rigorous science. Moreover, it's what makes many people who do care about science wonder if "modern medicine" is really first-and-foremost science-based any more. There seems to be an awful lot of armchair BS making it through this process.
- fabian2k 9y ago
  From a quick glance at the list of papers this seems to be mostly basic research. This makes it less likely that it could have affected treatment decisions. The formal peer review during publication is a very coarse filter, and not well suited to catch outright fraud. It's more of a quick sanity check than an in depth examination. The real verification happens once other scientists build on that paper and perform new experiments.
hueving 9y ago
It takes a special kind of sociopath to fake research on something with such potentially massive life/death consequences. This could have caused doctors to avoid a treatment; or, even worse in the grand scheme of things, prevented another researcher from entering into an area and making a big breakthrough.
- jakobegger 9y ago
  It's important to note that this is about fake reviews; not necessarily "fake research". There is an important difference. Fake research (eg. people fabricating measurements) often passes real peer review and is a different problem. The authors committing review fraud probably are not trying to publish fabricated research; it's more likely they were attempting to make it more likely that their research is published. Most articles are not rejected by reviewers because the research is "wrong"; rather because they are not significant enough, or other papers are better.
  - adolph 9y ago
    Where there is a whiff of smoke we often look for fire, no? Without a deep dive into a statistically significant amount of the research how could you say that those papers were probably not fabricated? It seems that part of the publishing apparatus discovered this flaw as the journal was undergoing a shift in publication models. In that all parties have an interest in the ongoing viability and prestige of the journal, and most often institutions respond to scandals by either coverup or diminishment, who benefits from the news and which party wants to contain the scandal to the review process only?
    - pizza_boy 9y ago
      There may be an intersection between fabricated research and fraudulently reviewed research but I understand that at least some of the papers retracted for reviewer fraud could have been published legitimately. Grandparent has it right. This is all about the pressure on authors to publish, and publish quickly.
- HarryHirsch 9y ago
  Cancer Biology is a journal with an impact factor of 3.1 or thereabouts, that's the equivalent of wrapping paper. People in the sciences know whose work is solid and whose isn't, with stuff like that you just go "Oh, it's XY again, with their usual rubbish", you don't even try to follow up. Broad & Wade called the phenomenon "the invisible boot". Required reading: https://www.amazon.com/-/dp/0192818899 https://www.amazon.com/-/dp/0192818899 A sociopath looks like the Sames-Sezen affair: http://blogs.sciencemag.org/pipeline/archives/2011/07/08/the_sames_sezen_fraud_case_holy_cow http://blogs.sciencemag.org/pipeline/archives/2011/07/08/the... Even there a certain blogger said that he knew "Dali Sames not to be a nice person", whoever went there knew exactly what he was getting into.
pvaldes 9y ago
Society put almost impossible obstables to scientists just for fun -> Society ends keeping a bigger percentage of liars than necessary, because who can lie convincingly can navigate the artificial walls and tests, and pass. If you require people with twelve real superpowers in your job offer, you will end hiring a liar in tighs.
kem 9y ago
I'm an editor at a journal, and these fake review cases have one thing in common, which is that the journal is using author suggestions for reviewers to solicit reviews. For those unfamiliar with the practice, this is common, and seems to becoming more standard at many journals, even (especially?) very reputable ones. I almost always ignore the reviewer's requests for particular authors, because it always seems to involve a conflict of interest. I am sensitive to requests for a particular author not to review a paper, because of vindictiveness in academics, but requesting reviewers is a different story. I've seen significant changes in this practice over time. Originally it seemed unusual, almost an option of last resort to cue the editor to bitter feuds or to help identify reviewers in areas that involve highly idiosyncratic issues. At a lot of journals this is still the case. Over time, though, it's seemed to turn into something different, a stopgap measure for overburdened, unconscientious, or harried editors to find increasingly scarce reviewers. A lot of attention has been paid to the pressures on researchers to publish or perish, quality be damned, and how that affects science. The flip side of that coin, though, is changes in editorial practices at many journals, which have become very cursory, rubber-stamping affairs. Turnaround time on papers is very quick at many journals, which gives authors quick feedback, but at the same time (not necessarily as a result) the review process has become shoddy. Editors are often unfamiliar with the subject material and don't take time to learn it, will grasp at reviewers who might be unqualified or have conflicts of interest, and treat reviews very superficially, even though studies have shown that they are horribly unreliable. To be fair, even finding reviewers can be difficult: I've had conversations with colleagues who have been told not to review because it doesn't bring in revenue to the department. Everyone wants to publish, but not everyone wants to review. Things are broken in science at the moment, at least in academics. Publishing through peer review has become extremely overvalued. I think it will maintain its place for certain reasons, but some change in mindset or culture around it is necessary.
- michaelhoffman 9y ago
  It's one thing to take author suggestions for peer reviewers--it's another thing altogether to use the contact details they supply also. Especially when they supply a gmail address for someone who should have a institutional affiliation. It's like calling the number left in a voicemail rather than the one on your bank's web site.
raverbashing 9y ago
Is this why we need "pay to publish" and also "pay to read"?
- sevensor 9y ago
  "Pay to publish" is something you don't see a lot of, outside of predatory journals. "Pay to read" does absolutely nothing to address this problem -- that money goes straight to the publisher and neither the reviewers nor the editor ever sees a dime. To reiterate, reviewers and editors are not paid in any way for their efforts.
  - new299 9y ago
    Pay to publish is not that uncommon e.g. https://academic.oup.com/nar/pages/Ms_Prep_Submission#charges https://academic.oup.com/nar/pages/Ms_Prep_Submission#charge... Sometimes this is under the guise of "open access" or other charges... there is often a waver if the author can not afford the fee (which you have to apply for).
  - rewrew 9y ago
    Why not pay the reviewers? I don't understand this -- they're the most important part of the process. Pay the reviewers for their time and that might start to turn this around.
fil_a_del_fee_a 9y ago
This may be downvoted to hell, but stuff like this makes me question things like vaccinating children. I am positive that hundreds of studies were conducted and "peer reviewed" to ensure that they were safe. And I am sure that the drug companies producing vaccines have 0 conflict of interest with the anti "fake-news" saying that "anti-vaxxers" are nuts. /s
- aggie 9y ago
  I don't know much about the science of vaccines so I won't address your point there. But I will say your comment is a good example of why I worry about the uptick in news of fraud and other bad practices in science, as well as the replication crisis is social psychology. These cases may be systemic or isolated or somewhere in the middle, but they erode public trust in good science. And with the 'post-truth' atmosphere of the politics, I fear a snowball effect.
- rsfern 9y ago
  I think you're being too quick to throw out the baby with the bath water. We know there are epidemiological risks to foregoing vaccination, dating to well before peer review and drug manufacturer incentives were problematic. If you're concerned about the incentives of vaccine manufacturers to produce safe products (which is a reasonable position), I think a better response in this case is to call for more transparency in research and manufacturing. Why accept a known large risk in an attempt to reduce a risk that may not exist at all?
- LeeHwang 9y ago
  Its not unreasonable. I really hope scientific authorities start realizing there is erosion of trust and they need to take extreme steps to fix it.
- jwilk 9y ago
  From the HN guidelines: Please don't bait other users by inviting them to downvote you or proclaim that you expect to get downvoted.
- ska 9y ago
  but stuff like this makes me question things like vaccinating children. Respectively, I suggest you question instead your understanding of how science works today (warts and all). This kind of thing is a problem, but it's not the same problem you seem to think it is.
culturalzero 9y ago
It's weird, and hate to be that guy, but is it significant that all of the people who did this appear to be Chinese? Springer lists all of the articles and authors.
- goodcanadian 9y ago
  Possibly. From the economist (http://www.economist.com/news/china/21586845-flawed-system-judging-research-leading-academic-fraud-looks-good-paper http://www.economist.com/news/china/21586845-flawed-system-j... [2013]): As China tries to take its seat at the top table of global academia, the criminal underworld has seized on a feature in its research system: the fact that research grants and promotions are awarded on the basis of the number of articles published, not on the quality of the original research. To be clear, there are plenty of high quality Chinese academics writing high quality papers, and I read somewhere that the Chinese government is well aware of the problem with fake journals and trying to address it, but it looks like the Chinese scientific community has even bigger problems of perverse incentives than elsewhere in the world.
netvarun 9y ago
For those wondering - here are the list of the papers being retracted: https://link.springer.com/article/10.1007/s13277-017-5487-6 https://link.springer.com/article/10.1007/s13277-017-5487-6

98 points by mathgenius 9y ago

Trust Science.

Wow, that's bonkers. I wonder if any of these papers has definitely convinced a decision-maker to pursue or avoid a particular course of treatment. IE, has someone literally died because of this? It seems likely, but these papers are only a year or two old. In any case, this kind of happening is why it's difficult anymore to see "peer review" as being tantamount to a verification that a conclusion was reached via rigorous science. Moreover, it's what makes many people who do care about science wonder if "modern medicine" is really first-and-foremost science-based any more. There seems to be an awful lot of armchair BS making it through this process.

fabian2k 9y ago
From a quick glance at the list of papers this seems to be mostly basic research. This makes it less likely that it could have affected treatment decisions. The formal peer review during publication is a very coarse filter, and not well suited to catch outright fraud. It's more of a quick sanity check than an in depth examination. The real verification happens once other scientists build on that paper and perform new experiments.

hueving 9y ago

It takes a special kind of sociopath to fake research on something with such potentially massive life/death consequences. This could have caused doctors to avoid a treatment; or, even worse in the grand scheme of things, prevented another researcher from entering into an area and making a big breakthrough.

jakobegger 9y ago
It's important to note that this is about fake reviews; not necessarily "fake research". There is an important difference. Fake research (eg. people fabricating measurements) often passes real peer review and is a different problem. The authors committing review fraud probably are not trying to publish fabricated research; it's more likely they were attempting to make it more likely that their research is published. Most articles are not rejected by reviewers because the research is "wrong"; rather because they are not significant enough, or other papers are better.
- adolph 9y ago
  Where there is a whiff of smoke we often look for fire, no? Without a deep dive into a statistically significant amount of the research how could you say that those papers were probably not fabricated? It seems that part of the publishing apparatus discovered this flaw as the journal was undergoing a shift in publication models. In that all parties have an interest in the ongoing viability and prestige of the journal, and most often institutions respond to scandals by either coverup or diminishment, who benefits from the news and which party wants to contain the scandal to the review process only?
  - pizza_boy 9y ago
    There may be an intersection between fabricated research and fraudulently reviewed research but I understand that at least some of the papers retracted for reviewer fraud could have been published legitimately. Grandparent has it right. This is all about the pressure on authors to publish, and publish quickly.
HarryHirsch 9y ago
Cancer Biology is a journal with an impact factor of 3.1 or thereabouts, that's the equivalent of wrapping paper. People in the sciences know whose work is solid and whose isn't, with stuff like that you just go "Oh, it's XY again, with their usual rubbish", you don't even try to follow up. Broad & Wade called the phenomenon "the invisible boot". Required reading: https://www.amazon.com/-/dp/0192818899 https://www.amazon.com/-/dp/0192818899 A sociopath looks like the Sames-Sezen affair: http://blogs.sciencemag.org/pipeline/archives/2011/07/08/the_sames_sezen_fraud_case_holy_cow http://blogs.sciencemag.org/pipeline/archives/2011/07/08/the... Even there a certain blogger said that he knew "Dali Sames not to be a nice person", whoever went there knew exactly what he was getting into.

pvaldes 9y ago

Society put almost impossible obstables to scientists just for fun -> Society ends keeping a bigger percentage of liars than necessary, because who can lie convincingly can navigate the artificial walls and tests, and pass. If you require people with twelve real superpowers in your job offer, you will end hiring a liar in tighs.

kem 9y ago

I'm an editor at a journal, and these fake review cases have one thing in common, which is that the journal is using author suggestions for reviewers to solicit reviews. For those unfamiliar with the practice, this is common, and seems to becoming more standard at many journals, even (especially?) very reputable ones. I almost always ignore the reviewer's requests for particular authors, because it always seems to involve a conflict of interest. I am sensitive to requests for a particular author not to review a paper, because of vindictiveness in academics, but requesting reviewers is a different story. I've seen significant changes in this practice over time. Originally it seemed unusual, almost an option of last resort to cue the editor to bitter feuds or to help identify reviewers in areas that involve highly idiosyncratic issues. At a lot of journals this is still the case. Over time, though, it's seemed to turn into something different, a stopgap measure for overburdened, unconscientious, or harried editors to find increasingly scarce reviewers. A lot of attention has been paid to the pressures on researchers to publish or perish, quality be damned, and how that affects science. The flip side of that coin, though, is changes in editorial practices at many journals, which have become very cursory, rubber-stamping affairs. Turnaround time on papers is very quick at many journals, which gives authors quick feedback, but at the same time (not necessarily as a result) the review process has become shoddy. Editors are often unfamiliar with the subject material and don't take time to learn it, will grasp at reviewers who might be unqualified or have conflicts of interest, and treat reviews very superficially, even though studies have shown that they are horribly unreliable. To be fair, even finding reviewers can be difficult: I've had conversations with colleagues who have been told not to review because it doesn't bring in revenue to the department. Everyone wants to publish, but not everyone wants to review. Things are broken in science at the moment, at least in academics. Publishing through peer review has become extremely overvalued. I think it will maintain its place for certain reasons, but some change in mindset or culture around it is necessary.

michaelhoffman 9y ago
It's one thing to take author suggestions for peer reviewers--it's another thing altogether to use the contact details they supply also. Especially when they supply a gmail address for someone who should have a institutional affiliation. It's like calling the number left in a voicemail rather than the one on your bank's web site.

raverbashing 9y ago

Is this why we need "pay to publish" and also "pay to read"?

sevensor 9y ago
"Pay to publish" is something you don't see a lot of, outside of predatory journals. "Pay to read" does absolutely nothing to address this problem -- that money goes straight to the publisher and neither the reviewers nor the editor ever sees a dime. To reiterate, reviewers and editors are not paid in any way for their efforts.
- new299 9y ago
  Pay to publish is not that uncommon e.g. https://academic.oup.com/nar/pages/Ms_Prep_Submission#charges https://academic.oup.com/nar/pages/Ms_Prep_Submission#charge... Sometimes this is under the guise of "open access" or other charges... there is often a waver if the author can not afford the fee (which you have to apply for).
- rewrew 9y ago
  Why not pay the reviewers? I don't understand this -- they're the most important part of the process. Pay the reviewers for their time and that might start to turn this around.

fil_a_del_fee_a 9y ago

This may be downvoted to hell, but stuff like this makes me question things like vaccinating children. I am positive that hundreds of studies were conducted and "peer reviewed" to ensure that they were safe. And I am sure that the drug companies producing vaccines have 0 conflict of interest with the anti "fake-news" saying that "anti-vaxxers" are nuts. /s

aggie 9y ago
I don't know much about the science of vaccines so I won't address your point there. But I will say your comment is a good example of why I worry about the uptick in news of fraud and other bad practices in science, as well as the replication crisis is social psychology. These cases may be systemic or isolated or somewhere in the middle, but they erode public trust in good science. And with the 'post-truth' atmosphere of the politics, I fear a snowball effect.
rsfern 9y ago
I think you're being too quick to throw out the baby with the bath water. We know there are epidemiological risks to foregoing vaccination, dating to well before peer review and drug manufacturer incentives were problematic. If you're concerned about the incentives of vaccine manufacturers to produce safe products (which is a reasonable position), I think a better response in this case is to call for more transparency in research and manufacturing. Why accept a known large risk in an attempt to reduce a risk that may not exist at all?
LeeHwang 9y ago
Its not unreasonable. I really hope scientific authorities start realizing there is erosion of trust and they need to take extreme steps to fix it.
jwilk 9y ago
From the HN guidelines: Please don't bait other users by inviting them to downvote you or proclaim that you expect to get downvoted.
ska 9y ago
but stuff like this makes me question things like vaccinating children. Respectively, I suggest you question instead your understanding of how science works today (warts and all). This kind of thing is a problem, but it's not the same problem you seem to think it is.

culturalzero 9y ago

It's weird, and hate to be that guy, but is it significant that all of the people who did this appear to be Chinese? Springer lists all of the articles and authors.

goodcanadian 9y ago
Possibly. From the economist (http://www.economist.com/news/china/21586845-flawed-system-judging-research-leading-academic-fraud-looks-good-paper http://www.economist.com/news/china/21586845-flawed-system-j... [2013]): As China tries to take its seat at the top table of global academia, the criminal underworld has seized on a feature in its research system: the fact that research grants and promotions are awarded on the basis of the number of articles published, not on the quality of the original research. To be clear, there are plenty of high quality Chinese academics writing high quality papers, and I read somewhere that the Chinese government is well aware of the problem with fake journals and trying to address it, but it looks like the Chinese scientific community has even bigger problems of perverse incentives than elsewhere in the world.

netvarun 9y ago

For those wondering - here are the list of the papers being retracted: https://link.springer.com/article/10.1007/s13277-017-5487-6 https://link.springer.com/article/10.1007/s13277-017-5487-6