Super fascinating. Do you have any links to this research?
https://scienceintegritydigest.com/about/
https://scienceintegritydigest.com/
She wrote an op-ed in the NYT last year: https://www.nytimes.com/interactive/2022/10/29/opinion/scien...
She says her work has led to 938 papers being retracted.
One of the interesting parts to me is that most of her work seems to be completely manual. She seems to have an eye for seeing photoshop on microscopy images or images of gels. The biggest method of fraud that she seems to find is researchers cutting and pasting from one image to another, for example, the scientist runs a gel but the results don't work out the way they hoped, so they paste in a line from another gel to make the experiment look like it worked.
The uncomfortable part is that this is just one of many ways that you could fabricate or alter research results, and Dr. Bik is only one person finding it in her spare time. Probably a whole lot of fraud goes undetected.
Brings to mind the skills of the 'super recogniser'. Perhaps her manual work involves similar aptitude.
https://en.wikipedia.org/wiki/Super_recogniser
--
Later edit: Indeed, the NYT article mentions -
"Since childhood, I’ve been “blessed” with what I’m told is a better-than-average ability to spot repeating patterns. It’s a questionable blessing when you’re focused more on the floor tiles than on the person you’re supposed to talk to. However, this ability, combined with my — what some might call obsessive — personality, helped me when hunting duplications in scientific images by eye." - Elisabeth Bik
> Most duplications were confirmed by ImageTwin.
> https://scienceintegritydigest.com/2023/06/27/concerns-about...
They are watching us.
(Gift link, no paywall)
After somebody who presumably works there posts a free link, you see the need to do this?
Both queries are reasonable ones to make. A spies by default without proper consent, B blocks by default without consent.
Is this optimal? Can we do better?
Moreover I recommend you have a slightly closer look if you are really holding up the nyt as a bastion of ethical behavior. But at least they stand up for Assange and journalism not being criminalised while doing stenography for the powerful (amongst some decent real reporting, some now being criminalized, away from the front page), right? I'm sure they'll balance their reporting about wikileaks by quoting someone who doesn't actively hate them any day now.
Anyway it should be noted that saying what you believe as you've done is the right thing and I endorse it even when defending the revenue a company that had $2.31 Billion on the books last year and has been a marker of inter-generational wealth and power for 170+ years to the family that owns it. I agree that what that looks like genuinely doesn't matter with regard to what you think is right in any ethical analysis. That goes unsaid too often imho.
Don't threaten me with a good time.
If j*urnos want money, they should choose to start acting in a way that deserves it at some point.
If j*urnos want money, they should choose to start acting in a way that deserves it at some point. In this case, if "The Paper of Record" needs to be regularly recorded from outside to be held to account for shoddy jobs and stitch-ups, exactly what service is it actually providing?
> Things could be about to get even worse. Artificial intelligence might help detect duplicated data in research, but it can also be used to generate fake data. It is easy nowadays to produce fabricated photos or videos of events that never happened, and A.I.-generated images might have already started to poison the scientific literature. As A.I. technology develops, it will become significantly harder to distinguish fake from real.
But reproducing studies could be live-streamed without a problem.
This could be done via software, and might catch more papers than the ~6k out of 100k that she did.
In the software world, there are tools for this, "software composition analysis". I worked at a company that got busted for violating GPL, and as part of settling the suit, all software had to be run through BlackDuck and all warning/issues found by the tool had to be resolved before the software could be released. (NOTE: the software that violated GPL was from an acquisition)