Show threadJeremyFeb 17

I mean, the irony is so goddamn deep.

You want to have a personal, authentic discussion with people about their careers and how that interfaces with resistance and social justice - and you have a fuckin AI chatbot that's taking notes and sharing it with who knows what. Because convenience.

SIGH.

I literally do not know why I need to keep having this discussion with adults when literal KIDS get it.

Museums, we don't have to be this way.

#museums #libraries #glam #RealEvalTalk #FuckAI

JeremyFeb 6

Yeah, this husk of the IMLS that remains can get lost with this re-alignment with bigotry.

It'll be years if not a full decade of recovery. So many people have been screwed this past year, and now we should take this "opportunity" to do further harm?

Nah. How's about no.

https://www.propublica.org/article/institute-of-museum-and-library-services-grant-guidelines-donald-trump

#IMLS #RealEvalTalk #museums #libraries #glam #funding #uspol

Grant Guidelines for Libraries and Museums Take “Chilling” Political Turn Under Trump

Former Institute of Museum and Library Services leaders from both political parties expressed concern that the new funding guidelines could encourage a more constrained or distorted view of American history.

ProPublica
Show threadJeremyJan 14

@nyhan Ugh. Yeah. Depends on who's going to read it, I suppose... And if there was a reasonable hope that the statement would bring about some change in process and data collection.

And at least two instances, I've had to include statements like that in reports because of the context and environment. I really didn't want to do it at first, but it was needed if I was going to honor the people I was working alongside. I have a feeling it wouldn't have been documented otherwise.

#RealEvalTalk

JeremyNov 18

There's a couple of things that stick out at me about the paper mentioned in this @404mediaco article:

https://www.404media.co/a-researcher-made-an-ai-that-completely-breaks-the-online-surveys-scientists-rely-on/

That sounds like a pretty big deal!

Besides the article being fairly hyperbolic - open recruitment online surveys aren't a tool that many social scientists rely on - this feels both a bit overblown and suspicious. Only one source was used in the 404 article - the research paper and its PI. So let's try and find the paper, right?

#research #evaluation #RealEvalTalk

A Researcher Made an AI That Completely Breaks the Online Surveys Scientists Rely On

We can no longer trust that survey responses are coming from real people.”

404 Media
Show threadJeremyNov 18

@DrSaucy @404mediaco Agreed, yet I'm sure we both have seen all kinds of work done under the banner of science that doesn't use any sort of validation or ACQ. With all the recent cuts to social science and data analysis work, I'm guessing there's going to be more people looking for shortcuts, so they won't do the things we know to do. It'll get pushed to an intern or put on the docket of someone without the necessary knowledge base.

#research #evaluation #RealEvalTalk

JeremyNov 10

I sometimes wonder if getting a PhD would turn me into a mediocre white man so that I could be reliably employed. Seems to work for a lot of people in non-profit and museum spaces.

Well, unless you're having to work with them. That right there sucks. So maybe I'll continue to skip the PhD.

#research #evaluation #RealEvalTalk #glam

Show threadJeremySep 30

@nyhan It's disheartening to be put in that position - like, this is where we are in the research and evaluation world??

And yet it's what needs to happen if they truly live into their values about equity. We're learning who is for real and who was performative.

#research #evaluation #equity #RealEvalTalk

Show threadJeremySep 11

Another big plus: the authors are trying to nudge folks into addressing the why and the how, instead of counting widgets. To me, that counts as major progress!

It also seems like libraries need to be hiring more evaluators. *cough *

#evaluation #research #libraries #RealEvalTalk

JeremySep 11

To me, the article does a good job outlining the problems for social science in LIS, sometimes unintentionally so. It seems very eurocentric (e.g. controlled trials being the epitome of research, what constitutes rigor, generalizability as the goal, etc). There are a fair bit of outmoded binary definitions, which is a bummer.

Another big plus: the authors seem like they're willing to interrogate their biases!

https://www.inthelibrarywiththeleadpipe.org/2015/ditchthesurvey-expanding-methodological-diversity-in-lis-research/

#research #evaluation #libraries #RealEvalTalk

#DitchTheSurvey: Expanding Methodological Diversity in LIS Research – In the Library with the Lead Pipe

JeremyAug 18, 2025

Really seems to be a broadly applicable lesson here from @jonny to so much around current flailing in nonprofit research and evaluation, but who am I to say.

https://neuromatch.social/@jonny/115051865161389984

#research #evaluation #nonprofit #RealEvalTalk

jonny (good kind) (@jonny@neuromatch.social)

you cloudbrained jokers have managed to convince everyone that object storage is some special thing when in reality it is just "normal http" and it "doesn't have folders" except "it totally does for every practical purpose except for the tooling doesn't let you use them like folders" and whatever makes it special on the backend doesn't matter to anyone except the people who administer object storage systems

Neuromatch Social