How Is the Surge in Fake Research Papers Impacting Science?

  • Thread starter jedishrfu
  • Start date
  • #1
14,824
9,200
https://www.theguardian.com/science...ers-push-research-credibility-to-crisis-point
Tens of thousands of bogus research papers are being published in journals in an international scandal that is worsening every year, scientists have warned. Medical research is being compromised, drug development hindered and promising academic research jeopardised thanks to a global wave of sham science that is sweeping laboratories and universities.

Last year the annual number of papers retracted by research journals topped 10,000 for the first time. Most analysts believe the figure is only the tip of an iceberg of scientific fraud.
 
  • Like
  • Informative
Likes BillTre and berkeman
Physics news on Phys.org
  • #2
Not really a surprise. It is the inevitable consequence of the introduction of papers per time as the only currency in science and justification for grants. Publish or perish - more than ever.
 
  • Like
Likes BillTre and Ibix
  • #3
The title and claims in the article are about scientific research in general, but all examples are from medical field specifically. Is it the field issue?
 
  • #4
Where there's money there will be fraud.

I suspect the medical field has the most money of any field because of its commercial possibilities.

I haven't heard of any math field papers although I'm sure there are some usually though they are in niche fields where there are few experts and consequently get ignored. However universities use papers as a gauge for tenure and that would be a major reason for some fraud.
 
  • Like
Likes BillTre
  • #5
I think that medical research is particularly at risk for producing questionable papers. This doesn't need to be on purpose. They might simply be less rigorous. I remember an instance many years ago, at a time when mainframes were the only available machines to deal with complex software and huge data. I wasn't on the phone but I knew the persons who operated support and helpdesk.

A student of medicine called the helpdesk and asked for support to use the statistics program on the mainframe. Of course, as so often, he had waited far too long before seeking help. We know this behavior well: "I need help, urgently!" is a frequent thread title before it is fixed. Anyway, the student had been asked about his sample size and it turned out to be a single liver of a mouse. He then had been told that this was way too few and that he needed more, not just one. Late at night he called again and said that he now had 10 samples. Being asked how he managed to get them so quickly, he responded that he used a scalpel to cut his liver in 10 pieces.

He might have become a good doctor, who knows, and he certainly received a Ph.D. but he most likely isn't a good scientist.
 
  • Wow
Likes berkeman and BillTre
  • #6
Medicine is likely the worst because there are many researchers and the confidence required is low. If you have 1000 papers with a 95% confidence level, 50 will be wrong. There is also the tendency to look for correlations anywhere you can. If I did the math right, if you look at N variables, the number of positive correlations by chance alone approaches N2/40.

N = 633 and you get 10.000 wrong results,
 
  • #7
Vanadium 50 said:
Medicine is likely the worst because there are many researchers and the confidence required is low. If you have 1000 papers with a 95% confidence level, 50 will be wrong. There is also the tendency to look for correlations anywhere you can. If I did the math right, if you look at N variables, the number of positive correlations by chance alone approaches N2/40.

N = 633 and you get 10.000 wrong results,
I think the social sciences such as psychology are probably just as bad as (if not far worse than) medicine.

I should also add that we need to make a distinction between bad science (e.g. where the experiments were poorly designed with insufficient sample sizes or effect sizes to provide meaningful measures of the effect being assessed, poorly thought out hypotheses, etc.) versus outright fraudulent sham papers (which is the topic of the Guardian article) where often times no research was done at all, and even the institutions and researchers may themselves be fake.

The distinction is important, because bad science may take place even where there is no intent on the part of the scientists involved to mislead -- often times, a combination of ignorance of statistics (or abusing statistics to search for any significant effects, often referred to as "p-hacking"), confirmation bias, and institutional pressures to publish novel findings can lead to bad science.
 
Last edited:
  • Like
Likes BillTre

Similar threads

Replies
7
Views
2K
Replies
2
Views
4K
Back
Top