The sale pitch for this AI Scientist "Kosmos" as presented in this Podcast just seems like a big HARKing exercise. Yes, just look at the data long enough and you'll find "something". Has anybody made any experiences with AI scientists models? Recommendations? pca.st/episode/58aa...

Agree with the quoted statement by @richarddmorey.bsky.social and disagree with Gelman on that point of asymmetry. Much of what we call metascience today has gained popularity and momentum from scientists/policy makers/funders is by making major positive claims about how science works/doesn't work.

When viewing the fake article in Google scholar on my university network, there is a link to access the article via my uni's library. That link sends me to a library page that makes fake article appear real... Turns out library page is made programmatically from info on Google scholar 🤦

The Max Planck Society has begun an exploratory round table for open science. We are drafting some recommendations to leadership. Still a long way to go! But here are my notes on the most recent draft, just so you all know how I am trying to steer things.

Google Scholar’s citation errors skew h-index leaderboards www.timeshighereducation.com/news/google-... "Thousands of mistakenly awarded citations left uncorrected highlight the perils of leaving profile curation to academics, say critics" I don't quite agree with the focus taken in this article 1/

we have 5 positions open at @goetheuni.bsky.social in Frankfurt - a 4,5 year postdoc in my @erc.europa.eu project on the educational cleavage - 2 postdoc & 2 PhD positions, 4 years, in a research group on Reconfiguration & Internalization of Social Structure www.uni-frankfurt.de/48794784/FB0...

Die Kürzungen in Hessen werden die #IchBinHanna & #IchBinReyhan unter uns leider besonders hart erwischen: Mittelbaustellen & Professuren werden nicht nachbesetzt und die 'Glücklichen' mit Stelle müssen deutlich mehr Arbeit schultern. Morgen vernetzen wir uns dagegen! Komm vorbei!⬇️ #nocutsinhessen

Did you know that from tomorrow, Qualtrics is offering synthetic panels (AI-generated participants)? Follow me down a rabbit hole I'm calling "doing science is tough and I'm so busy, can't we just make up participants?"

OSF | Qualitative Methods, and the Dangers of Positivism Creep osf.io/preprints/so... I think it is a valid concern that certain kinds of research amenable to #OpenScience practices crowd out alternative forms; the Gold Standard directive works in this direction www.whitehouse.gov/presidential... 1/

mehrere papieren gesehen habe, wäre schon damit geholfen, dass man einfach mal damit aufhört, die notwendigen publikationszahlen weiter zu erhöhen. ausserdem muss die betrachtung weg von "ich hab wieder was publiziert" hin zu "ich hab das und das rausgefunden". in meinem umfeld wird forschung als

There isn't a single problem "solved" by edtech that couldn't be fixed with smaller classes led by well-paid teachers given real academic freedom

Deadline Sunday! Submit applications for the 2026 Institute for Qualitative and Multi-Method Research (IQMR) to support early-career scholars in Arab #MENA countries. APSA and IQMR will support up to three MENA scholars to attend at the Maxwell School of Syracuse University. Apply here:

21. There was, however, one extremely positive new development, and that is the relatively new textbook by @saragoodman.bsky.social and Jennifer Cyr, a collaboration of 30+ female and non-binary scholars. ♥️ It is a breath of fresh air in the increasingly complex (U.S.) qualitative literature.

Increasing Precision of Terms Related to Reproducibility and Replicability www.cos.io/blog/increas... Any effort to systemize terms and their meaning is to be welcomed. The success in doing so depends on the ambition, which is not fully clear to me from the post. The linked preprint indicates 1/

“What I tell students is that, if what you’re trying to optimize publishing papers, you know, it’s actually honestly not that hard to do. Just do really crappy low-quality work and bomb conferences with it. But if you want to do really thoughtful, careful work, you’re at a disadvantage”

Probably a rhetorical questions, but I'll add the data anyway: We recently estimated—among other things—adoption of Bayesian analyses across different areas of psychology. The plots show the last 20 years and (rather simple) projections for the next 20 years. osf.io/preprints/ps...

"Sparzwang an Universitäten kommt im Hörsaal an. Viele Unis in Bayern haben Geldprobleme und müssen sparen. Studierende werden das jetzt zu Semesterbeginn auch spüren – obwohl die Staatsregierung sagt, dass noch nie so viel in Hochschulen investiert wurde wie im Moment." www.br.de/nachrichten/...

Ja: Alle, deren Verträge an der Uni Leipzig in den nächsten zwei Jahren auslaufen, sind raus. Die Lehre soll über Lehraufträge abgedeckt werden. Sachmittel werden gekürzt. Zudem sollen alleine an der Phil Fak 4 Professuren und 11 Mitarbeiterstellen abgebaut werden. Grund sind sinkende Zuweisungen.

The case for ending statistical significance reason.com/2025/12/01/o... There is much to like about estimation as an alternative, or maybe complement, to testing. I wonder whether a sweeping turn to estimation and effect sizes would imply putting up new goal posts with similar adverse incentives

Flensburg same: Sparmaßnahmen Wir befinden uns bereits in einer Phase, die die Uni als Ad-hoc Strategiephase bezeichnet, und das mindestens bis 2027 Alle Mittelbau Stellen werden nicht verlängert. Ausnahmen gibt es nur, wenn die Lehrbarkeit gefährdet ist..und de Uni so: "dafür gibt es Lehraufträge"

Academia is a daily gauntlet of criticism, rejection, & excessive workload demand. It's normal to feel like a failure when the main feedback is negative. And you're constantly doing a rushed job. That's why it's critical to have supportive managers. Alas, bullying & neglect are more common.

A blog post giving a more thorough take on survey experiments and the credibility revolution: cyrussamii.com?p=4168

A reflection on the credibility revolution in political science by @guygrossman.bsky.social et al: if you measure the influence of the credibility revolution by the relative proportion of articles using design-based tools, you may underestimate how impactful the credibility revolution has been!

This talk by @simine.com about what other fields can learn from the crisis in psychology is excellent. I so wish sociology was listening. If 10% of sociologists knew 50% of what she says then research in sociology would be 100% better.

Want to make nice graphs with me, starting next year? I'm hiring for a position at the University of Witten/Herdecke. uni-wh.softgarden.io/job/61280592...

Rearranging variables can make this clearer. In plot below, it's relatively easy to see that total effect needs no adjustment, and that direct effect from A to Y needs to block the indirect path that traverses D and C. D is a collider, and C the child of the collider, hence adjustment on B is needed

Readable, Reliable, Reusable: A Guide to Clean #rstats Code jacciz.github.io/portfolio/pr... Lots of good advice in this post. Too bad one (better, I) keeps forgetting about some practices in the process and follow bad practices

Periodically these stories break and we all agree it's wrong before going back to a system that rewards overwork, competition, gaming authorship, metrification and publishing for publishing's sake. Recognising these extremes as wrong it helps us recognise we're still participating - just a bit less

I am not opposed to editors publishing in their own journals, but look at the Management Science publication count for David Simchi-Levi, who was EIC from 2019 to 2023. YEAR. COUNT 1990 1 1992 1 2000 1 2019 2 2021 2 2022 10 2023 4 2024 1 2025 4

There are valid points in this post, but I disagree wrt key claims: - A significant result of a single study is not in itself conclusive. In order to know what one can learn from this, it would also be important to know how many tests of the same hypothesis did not reject it. 1/

Muller, T., & Adamson, J. L. (2025). Uncovering motivations behind authors’ questionable research practices: A collaborative autoethnographic investigation of editors’ experiences of academic publishing. Journal of Second Language Studies. Advance online publication. doi.org/10.1075/jsls...

I'm OK with our publication sequencing here (which isn't new): waiting for a response to be ready delays pub of the critique potentially quite a bit, which is also bad (& social media is full of angry threads about that practice - a mild example e.g. bsky.app/profile/ding...). That said (!): (1/2)

Biggest predictor of success was sample size, but largely due to ability to detect trivial effects. We need to realize that a non-quantitative hypothesis isn’t developed enough for hypothesis testing— if you can’t yet specify an expected effect size, keep working until you can, and *then* test.

An unbiased tally of social science experiments shows less than 1/3 of hypotheses were supported. This is what our research literature *should* look like. Biggest predictor of success was sample size, but largely due to ability to detect trivial effects.

Good morning. A coda (from me) on this replication business. While I understand the impulse, it is disheartening that most of the discussion generated by this replication is about the replication "process" -- both the APSR's process and my own process. What could we be talking about instead?

Encourage everyone interested in this exchange to read the authors' response and thread. I'll just offer two notes: 1. If you run many correlated tests, you unsurprisingly end up with many correlated answers. 2. There isn't a debate on inference - LPM with a binary DV has heteroskedasticity.

A critique of our (w/ @bertous.bsky.social) paper “Instrumentally inclusive” has just been published. Our response is under review (see below on process) but we feel obliged to share our draft for balance since the comment has been released without the response. osf.io/rn6h3/files/...

🔴 Live now: Webinar on the planned ERC Plus Grants. Join to learn how the scheme is being shaped to support outstanding principal investigators with bold ideas and a vision for transformative research. buff.ly/0VujxEM #ChooseEuropeforScience

"In 2019 we notified journals about serious integrity concerns in 172 clinical trials. Over five years later, only 22 have been retracted. The 135 unretracted trials have 1989 citations in systematic reviews, clinical guidelines, and consensus statements" [paraphrased] www.bmj.com/content/390/...

Zwischen Kaminabend und Mondmission Die DFG-Programmpauschale soll steigen, der deutsche ESA-Beitrag auch: Eine Woche der guten Nachrichten für Forschungsministerin Dorothee Bär. Im Wiarda-Blog: www.jmwiarda.de/blog/2025/11...

The reverse method is: open 12 papers, read the abstract, and feel intellectually empowered.

I recently heard from several journal editors that year-to-year submission rates are up by about 30%, sometimes more, sometimes less. Many papers have been described as being off-target for the journal at hand. It is not clear what the reasons are, but the odds are good that it is a mix of 1/

1. Pausing new submissions about AI topics for 90 days. That is, papers about AI models, testing AI models, proposing AI models, theories about the future of AI, etc. We will make exceptions for papers that are already accepted for publication (or published) in peer-reviewed scholarly journals /2

FYI: law and science, political science, and sociology all still have DDRIG programs run outside of the NSF but funded by NSF. The rest of these programs unfortunately do not. I run the political science DDRIG.

NEW from me - NSF cancels grant scheme for social science research. Seems the NSF quietly archived ALL calls for DDRIG grants in the SBE directorate. This is a massive blow for PhD students wanting to do cutting-edge social science research. 🏺🧪

ggskewboxplots: Enhanced Boxplots for Skewed Data in #Rstats arxiv.org/abs/2511.170... #dataviz Never thought about it, but asymmetric whiskers are plausible to me when the distribution is skewed. Plus I didn't know there are so many variants of boxplots (shouldn't be surprised, though)

METRICS is accepting applications for the 2026–27 postdoctoral fellowship in meta-research at Stanford. Deadline: Feb 15, 2026. Start date will be around Oct 1, 2026 (+/- 2 month flexibility). See: metrics.stanford.edu/postdoctoral... #MetaResearch #postdoc

Qualitative research with LLM chatbots: Technological reflexivity for interpretative technology journals.sagepub.com/doi/full/10.... I haven't tried it yet, but I think LLMs can be useful for some tasks for some kind of qualitative research. wrt the article, I am not sure how consistent the 1/

For some research studies the optimal sample size should be estimated at 0

🚨 LSE Assistant Professor in Political Science 🚨 We’re hiring a tenure-track assistant professor - any area of empirical political science - to join our wonderful Government Dept @lsegovernment.bsky.social Any questions, please reach out to me 📣 Please share! 📣 jobs.lse.ac.uk/Vacancies/W/...