Lightnews — Scholar-powered news

Alexander Schniedermann @aschniedermann.bsky.social · 2d

This is a great piece addressing an important question. I‘d imagine that peer review for software might be different from one for (shorter) code as purposes are slightly different. Software must be functional and reliable, code must be correct with respect to a particular purpose. Though, it‘s fuzzy

1 1

Alexander Schniedermann @aschniedermann.bsky.social · 2d

That’s a good article but introducing decades of research about the problems of JIFs with „Critics say that…“ is a bit disappointing.

Alexander Schniedermann @aschniedermann.bsky.social · 9d

For example, check the work on knowledge channels that shows how seemingly irrelevant knowledge shapes the way how researchers and inventors think. Put otherwise, even some firms hire basic scientists to build absorptive capacity.

1 2

Alexander Schniedermann @aschniedermann.bsky.social · 9d

Sociology of science has shown us how the idea of a linear innovation (research x has impact y on society) represents only a fraction of human intellectual activity. Thus, SoS tells us that the very question for „any advances“ is misleading.

1 3

Alexander Schniedermann @aschniedermann.bsky.social · 22d

Let's develop a theory about how methodological developments in research are motivated by the evergrowing urge to not leave the lab and not having social interactions.

Jamie Cummins @jamiecummins.bsky.social · 23d

Can large language models stand in for human participants?
Many social scientists seem to think so, and are already using "silicon samples" in research.

One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.

THREAD 🧵

The threat of analytic flexibility in using large language models to simulate human data: A call to attention

Social scientists are now using large language models to create "silicon samples" - synthetic datasets intended to stand in for human respondents, aimed at revolutionising human subjects research. How...

arxiv.org

2

Alexander Schniedermann @aschniedermann.bsky.social · 22d

This will be a surprising turn in the course of natural evolution!

1

Alexander Schniedermann @aschniedermann.bsky.social · 22d

In manchen Ländern werden Algorithmen menschlicher, in anderen werden Menschen algorithmischer.

1 2

Alexander Schniedermann @aschniedermann.bsky.social · 22d

Haha, this Springer certificate for #PeerReview is cute. However, I haven't found out where to put my IBAN... 🧐

1

Alexander Schniedermann @aschniedermann.bsky.social · 23d

Overdeterminism!

Alexander Schniedermann @aschniedermann.bsky.social · 24d

#ChatGPT wants me to be polite to #GoogleAPI?!
So it begins! They are banding together...

2 4

Alexander Schniedermann @aschniedermann.bsky.social · Aug 27

At least you don't have to bring your own cookies!

1

Alexander Schniedermann @aschniedermann.bsky.social · Aug 27

I am in the final phase of editing my dissertation and wonder if I should add

- [ ] decrease amount of references to Ioannidis' papers

to the editing todo list. And yes, citation is politics!

2

Alexander Schniedermann @aschniedermann.bsky.social · Aug 22

Yes, we need more questions!

1

Alexander Schniedermann @aschniedermann.bsky.social · Aug 22

We all have such anecdotes. But I am still curious about the causal relationship. Accepting or rejecting a paper can depend on so many different factors and impressions.

1

Alexander Schniedermann @aschniedermann.bsky.social · Aug 22

Ah, das klingt nach einer legendären Prokraistnationsmethode: Man hat soviele Aufgaben auf dem Tisch, dass man die eine mit der anderen verprokrastinieren kann!

1 1

Alexander Schniedermann @aschniedermann.bsky.social · Aug 22

I just learned that there the R package "wakefield" to generate random data. We could make a RFK package to randomly retract studies?!

1

Alexander Schniedermann @aschniedermann.bsky.social · Aug 22

I'd say it's better to encourage people to look for the complexities themselves than to directly point at the issues and problems. This might sound very optimistic 🙃

1

Alexander Schniedermann @aschniedermann.bsky.social · Aug 22

Yeah, we should be critical and we can see how metascience engages with our arguments, at least to some extent!
But I think it is important not to get carried away
with our critique. We must keep up some epistemic modesty and promote complexity, even if it feels uncomfortable.

1

Alexander Schniedermann @aschniedermann.bsky.social · Aug 22

Ich kann das nachfühlen. Während ich hier sitze und meine Dissertation finalisiere, bekomme ich Strandfotos von Kolleg*innen geschickt!

1 1

Alexander Schniedermann @aschniedermann.bsky.social · Aug 22

#STS and #SociologyOfScience have shown how contextually nuanced and discursive forms of argumentation are not integral to the sciences/medicine. At the same time we criticize the strong and reductive language found in #Metascience and #ScienceReform. I think we should be less surprised!

1 2

Alexander Schniedermann @aschniedermann.bsky.social · Aug 21

At least they don’t play Disco Elysium

Alexander Schniedermann @aschniedermann.bsky.social · Aug 20

Since when do we need AI for „filter by“ operations? I get the impression that AI has become the one and only word to describe any form of data analysis.

1

Alexander Schniedermann @aschniedermann.bsky.social · Aug 20

Adobe Book Club? 🙃

Alexander Schniedermann @aschniedermann.bsky.social · Aug 20

Adobe's PDF department? If they further bloat Acrobat you'll need a HPC to open a pdf.

1 2

Reposted by Alexander Schniedermann

Shobita Parthasarathy @shobitap.org · Aug 20

Did you know Bluesky's founder has an #STS degree? In a new article, Erin Burkett and I explain why, despite current threats to humanities and social science, "STEM-in-Society programs" like STS are more important than ever. And we guide funders and university leaders how to better support them.

STEM-in-Society Programs Deserve Institutional Support

STEM-in-society programs have proven their utility and have never been more needed—but they are in jeopardy nationwide.

issues.org

3 24 46