Academic publishers turn to AI software to catch bad scientists doctoring data

Shady scientists trying to publish bad research may want to think twice as academic publishers are increasingly using AI software to automatically spot signs of data tampering. Duplications of images, where the same picture of a cluster of cells, for example, is copied, flipped, rotated, shifted, or cropped is, unfortunately, …

[Notas Badoff]

But that's normal smell for code like this!

There are multiple places that track this issue. Retraction Watch is one such example and here's an example that specifically calls out image problems. "An institutional investigation has concluded that Figs. 1b and 2e are unreliable and that much of the additional data in the paper cannot be reliably verified from records."

Whether data problems result from sloppiness, time pressure, or simple dishonesty, research and academia *must* better their vigilance.

I could wish our industry had capable appraisers also.

[An_Old_Dog]

One Slide, Multiple Images

So wouldn't the data-forgers simply take the one slide they do have, take a pic, move the microscope stage so a different part of the (same) slide is under the lens, take a pic, and repeat as needed?

[Anonymous Coward]

Re: One Slide, Multiple Images

erm... it says in the text "[submages] may be shifted, flipped, or rotated; parts may be cropped, copied, or repeated"... so it's already checking for that, and more

[Anonymous Coward]

Academic publishers are using AI software to catch bad scientists doctoring data

while bad scientists doctoring data are using AI not to be caught.

[nautica]

The state of the scientific community

It's really sad that the state of the scientific community has come to this.

In all the days leading up to the mid-20th century, no scientist would have even considered trying any type of career-enhancing duplicity (there did exist outliers [outliars? I couldn't resist]; Blondlot, with his jingoistic "N-Ray" "discovery" immediately comes to mind).

The field was self-policing; any one who tried something like this would never again work in the field. Now, such activity is not only so commonplace that extensive and expensive measures are required to counter it; and it is often tried to parlay such efforts into 'book deals' and attempts at TV shows. Some 'scientists', unfortunately, put more stock in being a media star than in doing actual hard work...and, as has been said many times, "...science is hard."

[...this, from a research** scientist (most of whom refer to themselves simply, and very accurately, as "workers"), of many years.]

**"Research is what I’m doing when I don’t know what I’m doing."--Wernher von Braun

[Eclectic Man]

Re: The state of the scientific community

nautica: "It's really sad that the state of the scientific community has come to this."

Sir Cyril Burt (https://en.wikipedia.org/wiki/Cyril_Burt)

"Sir Cyril Lodowic Burt, FBA (3 March 1883 – 10 October 1971) was an English educational psychologist and geneticist who also made contributions to statistics. He is known for his studies on the heritability of IQ. Shortly after he died, his studies of inheritance of intelligence were discredited after evidence emerged indicating he had falsified research data, inventing correlations in separated twins which did not exist."

His flawed 'research' was primarily used by the educationalists to create the divisive 'Grammar School, and Secondary Modern School' system in the UK, so that the top 25% 'clever children' could go to Grammar Schools, and the rest would go to secondary modern schools. The segregation included the split between 'O'-Levels and CSE's (a top grade CSE was equivalent to an 'O'-Level pass) taken at 15 or 16. Because the rich wanted their children to go to good schools and not mix with poor kids from council estates.

We still have this educational apartheid in parts of England, and it is total hypocrisy, because the Grammar School head still want the best soccer, rugby, cricket and hockey players so offer them scholarships.

He also claimed to have analysed the intelligence of separated identical twins, when one was adopted by a poor shepherd, which would never have happened in 1950's Britain: it was only the rich middle and upper class who got to adopt healthy white babies

At least Sir Cyril did his fakery for a reason other than pure self aggrandisement. He was caught because his faked data contained far too many numerals 7, but his legacy lives on.

(Personal note, I went to a comprehensive school. One of my fellow students at York University want to a secondary modern school, got the best first in maths in our year, did part 3 at Cambridge, a PhD in York, got a fellowship to Oxford and is now a full professor at a prestigious university. So he at least beat the system.)

[Anonymous Coward]

From the other side of the "educational apartheid"

I was the first year in my county not to take the 11+. We all had to go to the new local Comprehensive (previously the Secondary Modern), which was a totally toxic place with only the least able students given any attention. I went from 3 years ahead when leaving Junior School, to just scraping enough O-levels to study for A-levels, but the school would not put on the ones I needed for my intended career - including that basic, Geography! And with there being no public transport in the countryside, my only option to pursue my career goals would have been to cycle nearly 8 miles each way, every day to the nearest Technical College (with no secure cycle racks).

When my older children went to school, (different school in the same county), Triple-Science (much like the older separate subjects of Biology, Chemistry and Physics) was the choice for all student unless they were unable. Now my youngest daughter has started GCSEs, Double-Science is the choice for all students and Triple-Science only for the lucky 25% (and it doesn't seem to have much relevance to ability! More related to time-tables.) When I wrote to the Headmaster to ask for a reconsideration of this policy, the reply was only semi-coherent, and I was so tempted to send it back covered in red ink indicating spelling and grammar issues!

Academies are measured on the metric of how many students get 3 GCSEs and so the work is dumbed down with that aim, rather than to stretch all students.

Basically, the education system seems to be failing all children to some extent, despite the stirling work by most teachers. But, most especially the more able - exactly the ones that the Grammar School system was meant to stretch.

[Anonymous Coward]

Re: From the other side of the "educational apartheid"

I was great when molly coddled but fell to pieces when I had to stand up for myself. Oh, poor poor me, it's all the systems fault. Instead of educating the "least able" they should have kept the poor in their place and allowed my parents to pay for me to be successful.

FTFY.

[AVR]

Re: From the other side of the "educational apartheid"

'red ink indicating spelling and grammar issues!'

'stirling work' - 'sterling work'

'the choice for all student' - 'the choice for all students'

I'm slightly dubious about your use of commas too.

Probably a good thing you didn't try to correct the headmaster.

[Cxwf]

Re: The state of the scientific community

>He was caught because his faked data contained far too many numerals 7, but his legacy lives on.

Well obviously, everyone knows 7 is the most random number!

[Coen Dijkgraaf]

I didn't know science was magic now

"The submages (sic) may be shifted, flipped, or rotated; parts may be cropped, copied, or repeated"

[Rich 2]

AI

Why is all new software these days “AI”?

[nautica]

Re: AI

"Why is all new software these days “AI”?"

Simple: real programmers have been replaced with, and managed by, Artificial Intelligentia.

[Atomic Duetto]

Re: AI

It’s just a brand, Algorithms Incorporated.

We live in the age of marketing after all

[nautica]

An immutable fact (very) conveniently forgotten...

"It doesn't matter how beautiful your theory is, it doesn't matter how smart you are. If it doesn't agree with experiment, it's wrong."

---Richard P. Feynman

[Anonymous Coward]

Re: An immutable fact (very) conveniently forgotten...

Who was paraphrasing Churchill's "However beautiful the strategy, you should occasionally look at the results."

[Auntie Dix]

Joel's "Uptown Dror Kolodkin-Gal"

"Proofig's...Dror Kolodkin-Gal..."

My screen reader made a choking sound and my spelling checker hoisted red flags.

I suspect that this article is fraudulent.