Keira (She/Her)

RE: https://mastodon.social/@kottke/116647849282379563

The thing I don't get here, is how?

I understand that AI hallucinates papers, it's very annoying and one of the reasons I don't use it for research.

But you're supposed to go read the paper before you cite it. Evaluate the paper, even. And you can't even skim or uncritically read just the abstract of a paper that doesn't exist.

So a huge number of published academics don't even read what they cite?

It's concerning that there's a higher standard asked of students than professionals.

#academic #academia #AI

May 27 at 9:24pmWeb
Show threadKeira (She/Her)May 27

I mean, I know people look at them uncritically. I was researching something for uni recently, and most of my source papers quoted the same paper regarding childhood malnutrition linked to adult diabetes.

So I went and read the paper. And found it doesn't say that at all. The results found what looked like a link, but the discussion clearly says they *couldn't* establish a link because there were far too many confounders. There was an older, larger study that legitimately did, but the pressure to cite more recent studies (and laziness?) meant most people were citing them anyway.

Show threadSu_GMay 27

@keira_reckons
Strange research that cites papers that don’t exist or papers that don’t contribute. 😐 Seems like a giant waste of time. Why is it rewarded at all. [I guess this is a rhetorical question; maybe a better question is: why do ‘research’ if this is all it is? I.e. where is the drive to fake it or do this fake research coming from?] 🤔

#researchIntegrity #research #citations #fakeResearch #reconnectingConsequencesToCauses

Show threadRaymondPierreL3May 28

@keira_reckons
Literature reviews are a fundamental part of any research and writing reseach papers. Given that almost all scientific/academic research is based on, or builds upon precedents. I don’t understand how we got to where we are today. Maybe the way #managerialism has corrupted many fine institutions with the ensuing #NeoLiberalFailures resulting in a lack of public funding and a scurrying for non-public funds (a potentially corrupting actor in any research) is the prime cause.

We;ve failed our #Scientific and #Academic Institutions and continue to fail them with each successive neoliberal govt.

#TaxTheRich #noBillionaires #PublicFinding #Education #Research #ResearchPapers

Show threadCarolannieMay 27
@keira_reckons Academics writing papers look at the cites used in papers by academics they like, or know, or they think has a good reputation. That's often good enough when you're under pressure to publish. Sadly. Been there....
Show threadKeira (She/Her)May 27

@carolannie That makes sense, and seems to be what happened in my nutrition example - all the serious people were citing the same source, so it must be true. Everyone cuts corners at work sometimes.

But the article says it's younger/newer academics, and at least the first one in any citation chain must have known they hadn't read the paper, and didn't see it cited - they had to have gotten it directly from an LLM and never tried to find it.

Show threadBennetMay 27
@keira_reckons @carolannie Yeah, even if I'm citing a paper on the basis of previous citations, I'll still go look up the original to enter it into my citation manager properly. Even that's enough to weed out fakes (I haven't found one yet), and it also gives me a chance to do a quick skim and check that it's at least relevant (and it's not uncommon for me to find that it's pretty marginal and go look up a better source).
Show threadthe input device is not a TTYMay 27

@keira_reckons

https://mastodon.me.uk/@pikesley/116450588361400655

Anti-Technology Extremist (@[email protected])

Remember also how little some people seem to care about quality and correctness

mastodon.me.uk
Show threadFabian TranschelMay 27

@keira_reckons It's more like:

a) Papers need references to be taken seriously. So you add references that sound like they could be related.

b) Scientifically rigorous scholars read the literature even *before* writing their papers. Like, during doing the research.

I keep repeating ad nauseam that false incentives lead to problematic (to say the least) practices.

And measuring scientific success with citations is the *central* problem in modern academia, even before grant practice.

Show threadKeira (She/Her)May 27

@ftranschel you're right.

I wonder how the authors of oft-miscited studies feel? Fine with all the citations racking up? Or annoyed that no one seems to have understood what they said?

Show threadKeira (She/Her)May 27

@ftranschel it does sometimes make me feel like a muppet for actually following the trails through.

I was explaining to a partner on a group assignment that he needed to go to the original paper, not cite a fact straight out of the introduction/background of some other paper that references it. He looked at me like I'd grown an extra head. (he wasn't even citing the referenced paper uncritically - he was citing the paper that did the referencing).

Show threadMarc A. Kastner6d ago

@ftranschel @keira_reckons for one recent conference I was asked to do a "LLM hallucination pre-check", between desk rejects and actual reviewer assignments basically checking all references of certain papers by hand and determine whether they exist or not.

I was kind of naively thinking it's just a formality, but then was pretty startled when I actually found one paper that had 4 or 5 completely made up references....

Show threadKeira (She/Her)6d ago
@mkasu @ftranschel oh wow, that's a lot
Show threadBen AvelingMay 27
It's both. People often don't read what they cite, and that's always been a problem. People know that everyone else cites such and such paper, so they also cite such and such paper.
But it's gotten worse, so much worse, with LLMs - people just trust the LLM to generate valid citations, and that's working as well as you'd expect.
@keira_reckons
Show threadRaymondPierreL3May 28
@keira_reckons
IMO, thinking off the wall, we should:
Kill the research paper publishing industry right off. Replace it with University-based publically funded peer-publishing freely available to the public. Restore trust in research papers, any otherwise published research could then be known to be biased.
#AcademicPublishing #ResearchPapers #RestoreTrust #FreeResearchPapers #ResearchIntegrity #SaveOurSciences