Tessera
Folgen

Martin Modrák

@modrakm.bsky.social

Biostatistics/bioinformatics at Charles University, 2nd faculty of Medicine. Bayesian in practice, but not a fan of Bayesian epistemology. Main on fedi: https://bayes.club/@modrak_m Blog: https://martinmodrak.cz

273 Posts1342 Follower46 Folgt
Joe Bak-Coleman@jbakcoleman.bsky.social · 5d ago

A must read by @jwlockhart.bsky.social on the money behind open science. He’s spot on that there are some subjects you just can’t touch with typical peer reviewers in #metascience scatter.wordpress.com/2026/07/31/s...

science funding and the cherry picking problem

Science publishing–like all of science–requires money. There’s no getting around it, even with the open science movement. What appears free to many authors and readers still requires money and infr…

scatter.wordpress.com

Julia M. Rohrer@dingdingpeng.the100.ci · 6d ago

New blog post! Odds ratios are...a bit weird. Here I explain two problems (interpretation, non-collapsibility) and provide some simple recommendations for improved reporting. www.the100.ci/2026/07/30/r...

Reviewer notes: Odds ratios are really odd

When psychological researchers investigate binary outcomes, they routinely report odds ratios to quantify the strength of an effect or an association. This is quite understandable given that psycholog...

the100.ci

Saloni@scientificdiscovery.dev · last wk.

I'm interested in finding and potentially supporting people who are interested in doing long-form writing on the big efforts in global health: Gavi, Global Fund, PEPFAR, etc. If this is you or someone you know, please make yourself known!

Aki Vehtari@avehtari.bsky.social · last wk.

priorsense R package has now JOSS paper "priorsense: Efficient prior and likelihood sensitivity checks for Bayesian models in R" you can cite, too doi.org/10.21105/jos... with Noa Kallioinen, Topi Paananen, and @paulbuerkner.com priorsense achieved also a gold badge from @ropensci.org review!

priorsense: Efficient prior and likelihood sensitivity checks for Bayesian models in R

Kallioinen et al., (2026). priorsense: Efficient prior and likelihood sensitivity checks for Bayesian models in R. Journal of Open Source Software, 11(123), 11036, https://doi.org/10.21105/joss.11036

doi.org

Martin Modrák@modrakm.bsky.social · last wk.

We just updated the preprint (w @paulbuerkner.com) to address some reviewer feedback. Most notably we added a case study where the model space is huge (2^100) and sampled explicitly with JAGS Things still work, though aggregating the results to get something meaningful becomes important. 1/2

Martin Modrák@modrakm.bsky.social · 5mo ago

Preprint alert: Simulation-based validation of Bayes Factor computation with @paulbuerkner.com and S. Stroppel. We bring lessons learned in SBC to validation of BFs. arxiv.org/abs/2508.11814 The idea is simple: simulate data from the models, fit and see if the inferences are calibrated. 1/10

Sanjay Srivastava@sanjaysrivastava.com · 2w ago

Imagine that in pre-publication peer review, the reliability of a 3-reviewer composite is .70. The journal accepts the top 10% of papers according to average review score. What is the reliability of the original reviewers in the subset of published papers? (Answer in alt text)

The reliability coefficient is -.77. No, the negative sign is not a typo.
AEA Data Editor@aeadata.bsky.social · 2w ago

Recent case, re-run of the authors' code failed on a first-stage regression, blocked code from completing. Hard to initially diagnose, because "qui do regression_55.do" shows no output... 🤦Back and forth with authors, they provide evidence that it works on their system. We send MWE to Stata support👇

Alejandro Sandoval-Lentisco@asandovall.bsky.social · 2w ago

How often are multiverse analyses implemented, not just cited/discussed? How are they implemented? 🧵 New preprint on the uptake and implementation of multiverse-style analyses 👇 www.biorxiv.org/content/10.6...

Uptake and Implementation of Multiverse-style Analyses Across 613 Studies

Empirical conclusions can depend on the many individual choices researchers make when analyzing data. Multiverse-style analyses address this by computing results across a set of specifications rather ...

biorxiv.org

Julia M. Rohrer@dingdingpeng.the100.ci · 3w ago

“In particular, by using a multiverse, a bad actor needs not commit to a single model and thus can raise the perceived bar for criticism of their position. Including bad or wrong model is not a problem for them as they are just asking questions and isn’t that what science is truly about?”

Martin Modrák@modrakm.bsky.social · 3w ago

Multiverse analysis, abdication of responsibility and manufacturing of doubt: I have written on some downsides I see with multiverse analysis (which I like in principle): arxiv.org/abs/2607.14623 I was inspired/provoked to write it by @dingdingpeng.the100.ci (thanks!)

Abstract: I argue that multiverse analysis is highly suited to two undesirable uses: abdication of researcher's responsibility for their conclusion and manufacturing of doubt. A review of multiverse analyses published in 2025 provides tentative empirical support that abdication of responsibility is present in the literature and I mention anecdotal evidence that multiverse has been used for manufacturing of doubt about Covid-19 precautions. To mitigate negative effects if multiverse analysis becomes widely used I suggest the community adopts two conventions for evaluating multiverse analyzes: evaluating multiverses by the single worst universe they contain and considering large size of a multiverse as a sign of weakness rather than a praiseworthy achievement.
Carl T. Bergstrom@carlbergstrom.com · 3w ago

The American Diabetes Association appears not to have learned their lesson the first time around (below). Now they've blocked publication of an editorial about this event that was slated for publication in the same journal, Diabetes Care, this week. But here it is. You know what to do.

Community Voices: On the Events of the American Diabetes Association 86th Scientific Sessions

The editorial and the seven opinion articles posted here were scheduled for publication on July 13, 2026, in the journal, Diabetes Care. All articles were provided, in advance of the scheduled publica...

zenodo.org

Carl T. Bergstrom@carlbergstrom.com · 2mo ago

From 500 views in the first two months to over 100,000 views in the last few days. The ADA opinion piece that was too scary to be distributed at the ADA meeting has gone full Streisand. Y'all are doing the Lord's work.

Download graph, topping at 103,071 total views as of a few minutes ago.
Berna Devezer@devezer.bsky.social · 3w ago

#Statsky I need your help. If you have a consulting gig or been involved in building a (stat. modeling) consulting center, would you be willing to share your business model, sample contracts/terms/conditions, your approach to recruiting clients... esp. in an academic setting, I'd appreciate it!

Ben Hanowell@hanowell.me · 4w ago

Hey, folks. I'm looking for open access person-time datasets that represent a competing risks structure. I want to analyze these datasets and write about those analyses to help people understand competing risks analysis methods. Nearly all of the data sets I have used for this are closed source.

Rafe Meager (they/them)@economeager.bsky.social · 4w ago

forever tapping the @protik.bsky.social sign: "you don't have to think someone is irredeemable to be able to notice that they haven't redeemed themselves"

Sunsette / Cynthia Skye@sunderedmarches.com · 4w ago

I can be an overly credulous believer in redemption and second chances. That is precisely why I never bought the redemption angle on Platner; he didn't actually do anything to redeem himself. He just kinda mumbled denials with annoyance between nondescript announcements that he'd changed.