Scientific Nonsense

[Chuckanut]

The next time you hear a news story about some scientific finding keep this article in mind. The author wrote a completely bogus paper that combined plagiarized quotes about geology and blood diseases. He included a photo of the Martian surface.

It was accepted for publishing by several journals if he paid. He didn't.

If nothing else, it's worth reading his mark-up of the article he 'wrote'. It's hilarious. I proably should have put this in the Thursday joke day thread.

Blinded by scientific gobbledygook

 

[travelover]

Sure, like we actually sent men to the moon.

 

[HFWR]

Good to be skeptical. There's a lot of stuff on the internet concerning grant money chasing hot topics, the pressure to publish, the peer-review process, etc.

 

[6miths]

Even pre-internet it was a big problem. We tell students that 90+% of what is in even 'prestigious' journals is not worth reading in the medical literature. All of the noise makes it more and more difficult to find the small number of studies that are of genuine value. The 'publish or perish' paradigm also inhibits the kind of research that needs to be done in medicine. Not many want to be the 51st author on the multi-center international trial that actually answers and important question when the university administrators don't see the investigator's role in that study as having any value. There was an excellent article in The Economist last year ago which pointed out many of the flaws in the way medical research is currently done.

 

[Midpack]

Good post. As others here have noted many times, always do your own research and your own numbers before accepting "expert" views.

The tricky ones aren't blatantly falsified, clever practitioners use selective data (knowingly omitting data that counters their claim), polling with biased questions, self-selecting source populations and all sorts of other tricks/spin. Whole TV networks do it every day, both 'sides of the aisle.' What "they" publish is technically correct, while deliberately and convincingly misleading.

 

[bld999]

"Simulations show that for most study designs and settings, it is more likely for a research claim to be false than true. Moreover, for many current scientific fields, claimed research findings may often be simply accurate measures of the prevailing bias."

John Ioannidis has done a lot of looking on this topic:

PLOS Medicine: Why Most Published Research Findings Are False

 

[growing_older]

There is a publicly available fake science paper generator available. People using it have gotten over 180 papers published. You can try it for yourself.

SCIgen - Wikipedia, the free encyclopedia

 

[CaliforniaMan]

I think this has always happened, and was my experience even in the 70s. This is one of the reasons I chose engineering rather than my first love, science, as my occupation. The proof of your being correct was that it actually worked, rather than dependent on some opinion of your colleagues/competitors.

I have many examples but one was we were looking for relationships between a large number of different species. We found some interesting relationships and wrote logical discussions about then. When I adjusted the acceptance level (95%) based on the number of different tests we tried, the significance disappeared, i.e. we got the same number of successes as expected due to chance. Nobody was interested in that, and the study went on.

In another example there was a large complex computer simulation I had written and run. I suspected it might have a flaw making the outcome incorrect, and reported it. But they said don't worry, we got the results we wanted, and the paper was published, and new grants received.

There were many other examples, but they all basically went like this, find something, publish it, find something else publish that, but don't publish too much in one paper or journal, spread it around to make it look better. So there were very few actually complete papers, just bits and pieces. And where I worked was one of the two top research establishments in the field in the U.S. In addition to this it seemed that the best researchers were often sidelined by the ones that had more political acumen or were working on subjects deemed politically more important (grantable). And this was in the 1970s!

 

[mickeyd]

Thanks for the post. As an armchair scientist, it's nice to see that so many others are as unknowing as am I about science matters.

 

[photoguy]

It was accepted for publishing by several journals if he paid. He didn't.

I'm having internet problems so I can't access the article and see what journals they were but I suspect this may be vanity publishing. I've never had to pay to get something published in a legit journal (however publishing standards do vary by field).

I also highly doubt that a nonsense paper would get accepted into any journal with a significant scientific reputation. In my experience a paper would get reviewed/read by at least three other individuals (editor and 2+ reviewers) and most reviewers (in my experience) would try to do a reasonable job (it's voluntary work anyway). Actually the editor should have filtered the paper out before even sending it to review.

Edit: I agree with many of the criticisms about the publishing process made in this thread, but these issues are somewhat different from gobblygook papers being accepted.

 

[scrinch]

I also highly doubt that a nonsense paper would get accepted into any journal with a significant scientific reputation. In my experience a paper would get reviewed/read by at least three other individuals (editor and 2+ reviewers) and most reviewers (in my experience) would try to do a reasonable job (it's voluntary work anyway). Actually the editor should have filtered the paper out before even sending it to review.

Edit: I agree with many of the criticisms about the publishing process made in this thread, but these issues are somewhat different from gobblygook papers being accepted.

Agreed. In my experience the peer review process works pretty well for reputable journals in my field (geology). We do have to remember that what is being published is usually data plus interpretation. In reading a journal article you expect the data to be correct and the interpretation to be reasonable (but not necessarily right). Science isn't black and white, and it doesn't usually progress by perfect step after perfect step. Issues get raised in journals, controversies develop, discussion/argument follows, and usually a consensus is developed over a period of months or years. The consensus doesn't always prove to be right, but generally progress is made. Don't expect THE TRUTH from a journal article.

On the other hand, if you submit a paper for publication and you hear back from the editor that it is accepted in less than a month, you can be pretty sure you chose a disreputable journal.

Sent from my Nexus 7 using Tapatalk

 

[AllDone]

Agreed. In my experience the peer review process works pretty well for reputable journals in my field (geology). We do have to remember that what is being published is usually data plus interpretation. In reading a journal article you expect the data to be correct and the interpretation to be reasonable (but not necessarily right). Science isn't black and white, and it doesn't usually progress by perfect step after perfect step. Issues get raised in journals, controversies develop, discussion/argument follows, and usually a consensus is developed over a period of months or years. The consensus doesn't always prove to be right, but generally progress is made. Don't expect THE TRUTH from a journal article.

On the other hand, if you submit a paper for publication and you hear back from the editor that it is accepted in less than a month, you can be pretty sure you chose a disreputable journal.

Sent from my Nexus 7 using Tapatalk

+1

"Publishing" in a vanity journal is not the same as publishing in a peer reviewed journal. Bad papers do get published in quality journals, but they get ferreted out eventually. That's how science works. Hypothesize, test, revise.

Disclosure: on an early date, my beloved showed me the reprints of his articles that had been published in "Science".

 

[scrinch]

+1
Disclosure: on an early date, my beloved showed me the reprints of his articles that had been published in "Science".

....and you said, "You had me at approved."

 

[LOL!]

Disclosure: on an early date, my beloved showed me the reprints of his articles that had been published in "Science".

I don't think I've ever told my spouse that my PhD thesis work was a cover article in "Nature".

Vanity publishing has been around awhile. Do those "Who's Who?" letters still end up in anyone's mailbox? How about getting your poems or short stories published? How about books on bike trips?

 

[ERhoosier]

"Simulations show that for most study designs and settings, it is more likely for a research claim to be false than true. Moreover, for many current scientific fields, claimed research findings may often be simply accurate measures of the prevailing bias."

John Ioannidis has done a lot of looking on this topic:

PLOS Medicine: Why Most Published Research Findings Are False

Ioannidis is correct to point out "prevailing bias" as a cause of false conclusions in research, but he largely ignores the troubling amount of outright fraud. And when this happens in medical research, there is potential for causing great harm.

Medical Research Fraud: Professors Go Unpunished in Glaxo $3 Billion Guilty Plea Over Paxil | The Nation

Dr. Sherri Tenpenny: Fraudulent Medical Research Could Affect Your Diagnosis

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3702092/

 

[AllDone]

Ioannidis is correct to point out "prevailing bias" as a cause of false conclusions in research, but he largely ignores the troubling amount of outright fraud. And when this happens in medical research, there is potential for causing great harm.

Medical Research Fraud: Professors Go Unpunished in Glaxo $3 Billion Guilty Plea Over Paxil | The Nation

Dr. Sherri Tenpenny: Fraudulent Medical Research Could Affect Your Diagnosis

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3702092/

Of course there is fraud. Look at the terrible suffering that Andrew Wakefield caused for a lousy $800K. There is a lot of money in medicine, particularly in pharmaceuticals. However, the fraud does get discovered. That's why you have to take the startling results with a grain of salt and wait for them to be replicated. Twenty years ago there was a paper that said that ulcers where caused by bacteria. Initially that raised a lot of eyebrows, but it turned out it was true. Now there have been a couple of small studies of antibiotics for the treatment of uncomplicated appendicitis. Maybe that will be the new approach, maybe not. That's why science operates by consensus.

 

[2B]

That's why science operates by consensus.

And even then, consensus frequently gets it wrong and has to be revised.

 

[robnplunder]

I get all my science facts from watching Star Trek. Who has time to read the papers