Who has a paper on why mediation analysis is bullshit / a legit thing we should do?

Physicist has written a fascinating big beautiful paper.Let’s not be afraid to call it what it is - groundbreaking. arxiv.org/abs/2603.21852

Where it all started. #Rstats

@chrisstaecker.bsky.social Apropos of today's ⚽️🏟🏆 match, I'm trying to make a trigonometry joke for the sequence Wirtz, Havertz, Haversine, Sine but I'm just going around in unit circles

We got a big bump in traffic yesterday when people learned about us on NPR. Thank you to @npr.org for covering our launch and bringing our story to people's attention! And thank you to all our new followers. We can do this together! 🌍️🧪💪 #ClimateAction #ClimateForUs #ProtectScience

A thought came across my head today. I wonder if there would be appetite for scheduled debates on Bluesky around specific topics related to causal inference / AI / TMLE/ stats where where some experts offer to comment / reply for a limited amount of time #askacadmia #episkty #statssky

Can't help but repost Gelman when he talks about Maine! 🫎 🦞 🫐

This paper is worth it for table 1 alone. 👏 @sabinehoffmann11.bsky.social @boulesteixlaure.bsky.social et al onlinelibrary.wiley.com/doi/10.1002/...

Excellent discussion Julia. @stephensenn.bsky.social would have a field day with this paper.

Had to read this one to figure out whether it's as click-bait-y as the title. As somebody who has written a lot about causal inference with observational data, I'm quite conflicted about it.> asteriskmag.substack.com/p/the-case-a...

Twin Piques is an absolutely diabolical name for this article by @stephensenn.bsky.social on the limitations/skepticism surrounding digital twins www.linkedin.com/pulse/twin-p...

Excellent new blog article by Steve Ruberg: analytixthinking.blog/2026/06/28/b... #Statistics #rct #pharma #clinicaltrial #StatsSky

#Statistics thought of the day: At the heart of much nonsense research lies data analysis (including machine learning) of a complexity that the effective sample size cannot support: hbiostat.org/hdata #StatsSky

@kdoudrick.bsky.social What a wild ride to read: #academicsky @elsevierconnect.bsky.social

Your reminder that fairy tales are full of stepmothers because women once routinely died in childbirth, from postpartum sepsis, or from other infectious diseases and left behind housefuls of motherless children. Stable nuclear families are made possible by vaccines, antibiotics, and safe abortions.

Remembering this WSJ graphic from 2007. Been thinking about "reverse statistics" lately, which means deriving the minimal set of assumptions that justify a result. Here I wonder if there is any function of the data that when max/min would produce such a curve. Ordinary Worst Squares?

R wins statistics award. statmodeling.stat.columbia.edu/2026/06/17/5...

Today I was reminded of a conversation with Terry Therneau at an R/Medicine conference some years ago (maybe 2019) where he told me a different version of the classic quote by George Box: “Remember that all models are wrong; the practical question is how wrong do they have to be to not be useful”

👏 Our congratulations to Emeritus Professor Brian Ripley on being named a 2026 Rousseeuw Prize laureate. The award recognises the contributions of Professor Ripley and the international R Core Team to statistical computing worldwide. Find out more at https://tinyurl.com/yjhcj4ja

I just keep thinking...why not RMST?

I feel like every time a new editor joins a major journal there should be a public hearing like we do with FDA commissioners. #academicsky #episky

I could not be happier to be a part of this wonderful group as a new co-organizer of the Boston R User Group! I’m excited to help restart this wonderful community along with @rachaeldempsey.bsky.social and @damiepak.bsky.social #rstats #statistics #rstats #boston

I think about this figure from @stephensenn.bsky.social's Statistical Issues in Drug Development a lot (Ch. 7) alot. A useful reminder that baseline adjustment is often about precision not just “controlling for confounding.” #StatsSky #Statistics #EpiSky

Honestly, Dawid and Skene's 1979 paper is one of the most captivating and well-written statistics papers I've ever read. It's remarkable that, nearly 50 years later, people are still discussing it, building on its ideas, and even finding inspiration in it to pursue statistics themselves.

The Kappa Zoo: David Eubanks’s online monograph on rating models statmodeling.stat.columbia.edu/2026/05/28/t...

The basis of science is to have reproducible work. I would say the methods and results are the most important - everything else is just context

I was literally trying to find better ways to plot multivariate distributions a few weeks ago - outstanding

Announcing #RStats package ggcube, now on CRAN -- make 3D #ggplot2 figures using `coord_3d()` matthewkling.github.io/ggcube/

Okay now people are really making up symbols:

*** NEW PAPER *** The CRT-Estimands Framework - a consensus based extension of the ICH E9(R1) addendum for cluster randomised trials - now been published in the @bmj.com. -> www.bmj.com/content/393/... + commentary in @jama.com -> jamanetwork.com/journals/jam... @unibirmingham.bsky.social

#Statistics thought of the day: Virtually all of single-event survival analysis (including Kaplan-Meier, Cox PH, AFT) is a special case of cumulative probability semiparametric ordinal regression models: hbiostat.org/rmsc/ordsurv #StatsSky

Very Excited for the next Causal Inference Seminar at @jhubiostat.bsky.social on May 21st! #statsky #academicsky #skystats

Came across this gem again today. Still makes me giggle @richarddriley.bsky.social @gscollins.bsky.social Yes, it is published: pubmed.ncbi.nlm.nih.gov/32817260/

Wonderful manuscript on the need and power of Statistical Methodology Groups in Pharma , authored by Jennifer Davenport at Roche, @numbersman77.bsky.social, and others! Below are some of my favorite quotes #skystats #statssky #statistics #EpiSky

Great to see this investigated. I suspect we'd see it everywhere if we bothered to look. More thoughts below, in the context of diabetes. www.thelancet.com/journals/lan... (with @f2harrell.bsky.social and @maartenvsmeden.bsky.social)

"Conventional statistical practice, which systematically ignores certain sources of error ..., is bound to yield overconfident results that will often fail to replicate." New preprint by Knief & Forstmeier (2026): doi.org/10.1101/2025... #stats #MetaSci

I absolutely love this. Caregiver grants for those who need it to join @amstatnews.bsky.social JSM this year. I hope more conferences move in this direction. #academiceky #skystats #statssky #JSM2026

Starting in 5 minutes @jhubiostat.bsky.social Zoom link: publichealth.jhu.edu/events/2026/...

glmmtmb has always been my go-to, would love to see how this compares

This is the clearest and most accessible introduction to DML I've ever read: arxiv.org/abs/2504.08324 Congrats to the authors @aahrens.bsky.social, @markeschaffer.bsky.social, et al on this amazing paper and great accompanying R package! I'm now DML-pilled.

Stunning! Definitely, thanks for sharing.

I did not know this.

"The algorithm did not become compassionate; however, the medical profession has drifted so far from the bedside that an AI language model can now outperform physicians on the very quality that medicine assumed would remain uniquely human." Great viewpoint @jama.com #skystats tinyurl.com/4yv8b6y5

Regression Modeling Strategies 4-day virtual course May 14-15, 18-19. More than regression:principled flexible modeling, accuracy, interpretation, selection, stability, validation, Bayes, contrast w/ML &more. hbiostat.org/course @amstatnews.bsky.social @instats.bsky.social #Statistics #StatsSky

Just learned that Oscar Sheynin died a couple of years ago. I always enjoyed his pieces on the history of statistics. As an associate editor for Biometrika in the min 1990s I had no hesitation in recommending this academic.oup.com/biomet/artic... coauthored with Pfanzagl for publication.

8 workers on a 1 GB dataset no longer means 8 GB of RAM - it means 1 GB, shared. Massive step forward for simulation studies. #statssky #slystats

Thrilled to realize my JHU schedule overlaps with @pwgtennant.bsky.social for my first in-person causal inference seminar! @jhubiostat.bsky.social

Estimating prevalence is really really really difficult

I’m so glad to see more attention being paid to the data being used to train AI models. The results will only be as good as the data going in.

This might not seem like a big win, but I learned how to make custom .csl files in Zotero today #academicsky