iiinova — where your digital self keeps evolving

Reposted by Magnus Johansson

Tomas Formo Langkaas @tflangkaas.bsky.social · 29/09/2026

“Whether idionomically guided treatment does better is itself a causal question about groups, answerable by randomizing patients to idionomically guided or standard treatment.” Well put. (Keep reading “idiotically” though)

031

Reposted by Magnus Johansson

Ian Hussey @ianhussey.mmmdata.io · 29/09/2026

Hayes has been banging this drum about idionomic analyses for 6+ years, with folk like @jamiecummins.bsky.social pointing out that its reinventing-the-wheel-but-worse for just as long. My experience is that the ACT community is insufficiently interested and aware of existing methods+stats work.

12311

Magnus Johansson @pgmj.bsky.social · 28/09/2026

Our commentary "A Revolution in Search of an Estimand: On the Idionomic Proposal of Hayes et al. (2026)" is now available as a preprint: osf.io/preprints/ps... With @rpsychologist.com, @matti.vuorre.com, and Nils Hentati Isacsson.

osf.io

OSF

03118

Magnus Johansson @pgmj.bsky.social · 25/09/2026

We just published a meta-analysis, where one finding is that "number of sickdays" is a bounded, not seldom zero/one-inflated variable, which is unsuitable for meta-analysis based on mean/SD. Today, I came across a meta-analysis including sickday stats like mean=9 and SD=27. doi.org/10.1186/s403...

doi.org

Client Challenge

131

Magnus Johansson @pgmj.bsky.social · 24/09/2026

This is worth reposting once in a while. bsky.app/profile/rpsy...

0145

Reposted by Magnus Johansson

Canadian Paintings @canadianpaintings.bsky.social · 23/09/2026

Rounding the Horn Terry Watkinson 2020

019649

Magnus Johansson @pgmj.bsky.social · 23/09/2026

Version 3.2.2 of my jamovi module for #Rasch #psychometrics is now available. It brings conditional reliability, a Reliable Change Index with exact critical values, and a new targeting figure. The module uses #rstats pkg easyRasch2 v1.3.1 and produces identical results. pgmj.github.io/easyRasch2/

0101

Magnus Johansson @pgmj.bsky.social · 20/09/2026

Makina 67 (orange filter) Rollei RPX400@800

Black and white photo of a sunlit oak tree.

050

Reposted by Magnus Johansson

Nicolas Hénin @nicolashenin.net · 17/09/2026

Nice campaign by the Norwegian Consumer Council about the declining lifespan of products. Enjoy!

180140426455

Reposted by Magnus Johansson

Wes Bonifay @wesbonifay.bsky.social · 15/09/2026

It's right there in the title europepmc.org/article/med/...

4173

Magnus Johansson @pgmj.bsky.social · 15/09/2026

It only recently became apparent to me that some psychologists publish "method development" in applied journals, which means reviewers are not properly equipped. Looking at the where this group has published on this topic before makes the pattern clear in this case at least. Why did CPR invite this?

120

Reposted by Magnus Johansson

Robert (Bob) Kubinec @rmkubinec.bsky.social · 15/09/2026

I see this comment a lot: I shouldn't use Bayesian models without a good reason. Frequentist packages (maximum likelihood, etc) are a special case of Bayesian analysis. Bayes can do *any* MLE model plus a lot more. So the question is, what is your justification for restricting yourself to MLE?

3152

Magnus Johansson @pgmj.bsky.social · 14/09/2026

Let's say I scrambled together a comment/critique of this paper. Any suggestions on journal that I should submit it to?

420

Reposted by Magnus Johansson

Ade Taylor (shoot_dev_print_repeat) on instagram @ade-pano.bsky.social · 12/09/2026

Pano triple Saturday Hasselblad Xpan ii Ilford FP4

6433

Reposted by Magnus Johansson

Solomon Kurz @solomonkurz.bsky.social · 04/07/2026

At a basic level, I think this depends on what you care about. If you want quick pubs for a grade or a promotion, just do t-tests. If you are interested in your data, learn how to fit and interpret ordinal models.

66114

Magnus Johansson @pgmj.bsky.social · 09/09/2026

Curiously, I again encountered a CFA paper that refers to Marsh et al 2004 to justify rule-of-thumb cutoffs to interpret model fit. Marsh's title is "Comment on hypothesis-testing approaches to setting cutoff values for fit indexes and dangers in overgeneralizing Hu and Bentler’s (1999) findings"

331

Reposted by Magnus Johansson

Vincent Arel-Bundock @vincentab.bsky.social · 04/09/2026

Big news! 🎉 𝚖𝚊𝚛𝚐𝚒𝚗𝚊𝚕𝚎𝚏𝚏𝚎𝚌𝚝𝚜 1.0.0 for #Rstats is out. It’s a big number and it feels like a big step. I wrote a blog on the challenges of interpreting statistical models, SPEED, cool new features, the future, and a 5 year package development and writing odyssey. arelbundock.com/posts/margin...

1642197

Reposted by Magnus Johansson

Aki Vehtari @avehtari.bsky.social · 26/08/2026

New ppc_calibration() plots for posterior predictive calibration checks for binary outcome models. Previously we advocated a bootstrap variant. Now new posterior and loo variants by @teemusailynoja.bsky.social and @florencebockting.bsky.social as described in teemusailynoja.github.io/visual-predi...

bayesplot::ppc_calibration() plots for Negative-binomial and Zero-inflated negative binomial models

171

Reposted by Magnus Johansson

MC Stan @mc-stan.org · 18/08/2026

There is now StanCon 2026 Uppsala playlist www.youtube.com/playlist?lis... At the moment talks by @mjskay.com, @aseyboldt.bsky.social, @paulbuerkner.com, @charlesm993.bsky.social are up, and we'll keep uploading more during the week

youtube.com

StanCon 2026 Uppsala - YouTube

23717

Reposted by Magnus Johansson

Ian Hussey @ianhussey.mmmdata.io · 14/08/2026

I’ve been told for years that it’s legally and ethically impossible to share medical RCT data. In 2024, the BMJ made sharing anonymised data mandatory, and now lots* of folk do it. Interesting how the barriers evaporated like that, isn’t it? *not everyone, compliance problems exist.

medrxiv.org

Availability of clinical trial individual patient data in the BMJ before and after adoption of a stringent data-sharing policy, compared with recent rates at other major medical journals

Background In 2024, the BMJ updated its data-sharing policy for clinical trials, requiring open deposit of deidentified individual participant data (IPD) before publication. We considered whether data...

510040

Magnus Johansson @pgmj.bsky.social · 10/08/2026

Helpful cat also likes the vertical mouse.

010

Reposted by Magnus Johansson

Noah Greifer @noahgreifer.bsky.social · 03/08/2026

I'm so happy to announce version 2.0.0 of my #Rstats package WeightIt is out on CRAN! New features: censoring weights, multilevel propensity scores, improved weights for continuous treatments, bias-reduced ordinal and multinomial models, M-estimation in subgroups Check out the website below!

ngreifer.github.io

Weighting for Covariate Balance in Observational Studies

Generates balancing weights for causal effect estimation in observational studies with binary, multi-category, or continuous point or longitudinal treatments by easing and extending the functionality ...

29037

Reposted by Magnus Johansson

R Works @rworks.bsky.social · 29/07/2026

An astounding 412 (!) #RStats packages made their way to CRAN last month, and @revojoe.bsky.social picked out forty that he thinks you should know about. Check them out on R Works: rworks.dev/posts/june-2...

rworks.dev

June 2026 Top 40 New CRAN Packages – R Works

A personal, idiosyncratic, unabashedly biased, time-constrained attempt to capture the depth and breadth of the new packages submitted to CRAN in a single month.

052

Reposted by Magnus Johansson

Aki Vehtari @avehtari.bsky.social · 29/07/2026

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!

doi.org

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

16816

Reposted by Magnus Johansson

Jamie Cummins @jamiecummins.bsky.social · 14/07/2026

Today we launch the first stable release of RegCheck: v1.0.0. RegCheck makes it easier and quicker to compare study registrations to published papers for consistency - something we know is important in principle, but rarely done in practice. A 🧵 on what's new: regcheck.app

regcheck.app

RegCheck

RegCheck is an AI tool to compare preregistrations with papers instantly.

512783

Magnus Johansson @pgmj.bsky.social · 10/07/2026

Two papers accepted today! Unusual coincidence, but also have a lot of papers under review currently.

060

Reposted by Magnus Johansson

Petter Törnberg @pettertornberg.com · 07/07/2026

‘Polarization’ has become the master concept for diagnosing our current democratic crisis. But this is not polarization. It is the radicalization of the political right. By misdiagnosing it, we risk becoming complicit in the crisis. 🧵 Just out in Perspectives on Politics doi.org/10.1017/S153...

4834451428

Reposted by Magnus Johansson

Jamie Cummins @jamiecummins.bsky.social · 08/07/2026

WE ARE HIRING! Come work with us (me + @malte.the100.ci) on developing and validating RegCheck. PhD or postdoc applications welcome. Folk with backgrounds in meta-science, psychology, (pre)clinical trials, and/or economics are especially invited to apply. jobs.unibe.ch/job-vacancie...

jobs.unibe.ch

PhD Student/ Postdoc Position in Meta-Science - Universität Bern

Universität Bern is looking for PhD Student/ Postdoc Position in Meta-Science

15648

Reposted by Magnus Johansson

Dr. Jonathan Foley @globalecoguy.bsky.social · 08/07/2026

We are excited to relaunch our Drawdown Learn program, offering free online courses designed for climate professionals and lay audiences. Currently, we have 3 courses in production: - Living on Planet Earth - Getting Started with Climate Solutions - Exploring Food & Climate Solutions

33911

Magnus Johansson @pgmj.bsky.social · 02/07/2026

Another example of how useful {geomtextpath} can be, imho. This is from the item category probability plot in my {easyRasch2} package. Most psychometric papers within the factor analytic tradition seem to disregard evaluating how the response scale works, which I find difficult to understand.

6629

Reposted by Magnus Johansson

Kristoffer Magnusson @rpsychologist.com · 25/06/2026

New interactive blog! "Why Adjusted Regression Coefficients Are Less Descriptive Than They Look" rpsychologist.com/descriptive-...

1520769

Magnus Johansson @pgmj.bsky.social · 25/06/2026

I tried to update a preprint at OSF.io 2 weeks ago, but at the last step I realized I needed to make an edit and pressed cancel. Then the preprint was deleted from OSF. Support quickly responded that this is a known bug and they will restore the preprint, but this has yet to happen :(

200

Reposted by Magnus Johansson

Values to Action @v2action.bsky.social · 23/06/2026

🎙️ Our newest podcast episode features Dr. Diana Fishbein discussing prevention science, child wellbeing, and why America needs a "Decade of the Child." Listen and subscribe: snapt.io/EGFOy 💬 What would make the biggest difference for children in your community? #PreventionScience #ChildWellbeing

022

Reposted by Magnus Johansson

Robert (Bob) Kubinec @rmkubinec.bsky.social · 23/06/2026

If you would have also said yes, then there is something missing in your understanding of causal inference. RCTs *do not* provide information about what would have happened to a person if they had received the treatment. They provide information about average *group-level* differences.

1214

Reposted by Magnus Johansson

Wolfgang Viechtbauer @wviechtb.bsky.social · 22/06/2026

The standard random-effects model in meta-analysis assumes that the amount of heterogeneity is homoscedastic. This assumption may not be true. I derived and examined various tests for heteroscedastic heterogeneity. See osf.io/cavhy for a preprint describing these methods. #MetaAnalysis #RStats

osf.io

OSF

13112

Reposted by Magnus Johansson

Vincent Arel-Bundock @vincentab.bsky.social · 20/06/2026

🎉 Announcing Calepin 🎉 A new tool to create computational notebooks and static websites, built on Typst. Write code and prose directly in your .typ files, then publish them as documents or as a full website. vincentarelbundock.github.io/calepin/ 1/6 🧵 #Rstats #Pydata #Typst

611026

Reposted by Magnus Johansson

John Altringham @johnaltringham.bsky.social · 16/06/2026

Watercolour sketches started yesterday and finished this morning. Changing light over the Sound of Sleat from Isleornsay, Skye. #watercolour #landscape #mountains #Scotland #art

Watercolour in blues. Early mist and cloud are clearing from mountains seen across a sea passage. The sea is deep blue under a band of heavy cloud but for a bright, sunlit patch in the centre.

Watercolour in blues. Low mist obscures the meeting of mountains and sea across the centre of the picture. The sea is almost calm and reflects the mountains and sky.

5518

Reposted by Magnus Johansson

Mike Frank @mcxfrank.bsky.social · 15/06/2026

This week's Experimentology chapter is "Estimation". The most important move in planning an analysis isn't picking your test or your model — it's clearly defining what you're trying to estimate. 🧵 experimentology.io/005-estimat...

1177

Magnus Johansson @pgmj.bsky.social · 15/06/2026

I had a great time at the Scandinavian Applied Measurement Conference in Copenhagen last week! On the last day, I presented on the topic of "Assessing model and item fit in psychometrics: Moving toward up-to-date methods and appropriate cutoff values". Slides available here: pgmj.github.io/SAMC2026/

pgmj.github.io

Assessing model and item fit in psychometrics

120

Reposted by Magnus Johansson

Protzko @protzko.bsky.social · 14/06/2026

#psych content journals need sections for pure methods papers. Methods papers are put in methods journals read by methods people. But non-methods folk don't read those journals, so they miss the methodological insights/updates. Where are the content journals that eagerly publish methods papers?

1133

Magnus Johansson @pgmj.bsky.social · 08/06/2026

Been reworking my `easyRasch` package for CRAN, and it is finally available. This is also the code base for my jamovi module easyRasch2jmv. pgmj.github.io/easyRasch2/ #psychometrics #rstats #openscience

151

Reposted by Magnus Johansson

Måns Magnusson @mansmag.bsky.social · 03/06/2026

🔥 The early-bird deadline for StanCon 2026 (and, even more importantly, the StanCon Uppsala T-shirt) is in one week! 🔥 More information can be found here: discourse.mc-stan.org/t/stancon-20...

discourse.mc-stan.org

StanCon 2026: Early bird (and t-shirt) deadline in one week!

Hi all! Also, the 10th of June (AoE) is the early-bird deadline, and after that, the conference fee will increase! Also, only those who register before the 10th of June will be able to buy a StanCon ...

174

Reposted by Magnus Johansson

Samuel Moore @samuelmoore.org · 03/06/2026

"The editorial team’s resignation follows ongoing concerns about Springer Nature putting pressure on editors to accept and publish more articles. Commercial publishers like Springer Nature are increasingly using this tactic to grow the revenue they receive for open access publishing via APCs."

openlibhums.org

New Open Access Journal to Launch Following Resignation of Editorial Team from <em>Natural Language Semantics</em>

The Open Library of Humanities is pleased to announce the launch of a new diamond open access journal in linguistics, following the resignation of the editorial team of Natural Language …

0198

Reposted by Magnus Johansson

Eiko Fried @eikofried.bsky.social · 02/06/2026

I wish I could write like this. And I wish Lindy West wrote tutorial papers in psychometrics. (Source: Lindy West, 2019 essay collection 'The Witches Are Coming')

742313326

Reposted by Magnus Johansson

Francisco Rodriguez-Sanchez @frodsan.bsky.social · 31/05/2026

{easystats} is not only highly functional and powerful, but their help is also a goldmine of accessible #statistics advice. Here explaining how to address high VIF values (hint: generally not by removing predictors, as often done) easystats.github.io/performance/... Thanks @easystats.github.io

High VIF values indicate that coefficient estimates may be unstable and have inflated standard errors. However, removing predictors with high VIF values is generally not recommended as a blanket solution (Vanhove 2021; Morrissey and Ruxton 2018; Gregorich et al. 2021). Multicollinearity is primarily a concern for interpretation of individual coefficients, not for the model’s overall predictive performance or for drawing inferences about the combined effects of correlated predictors.

Consider these points when dealing with multicollinearity:

If your goal is prediction, multicollinearity is typically not a problem. The model can still make accurate predictions even when predictors are highly correlated (Feng et al. 2019; Graham 2003).

If your goal is to interpret individual coefficients, high VIF values signal that you should be cautious. The coefficients represent the effect of each predictor while holding all others constant, which may not be meaningful when predictors are strongly related. In such cases, consider:
Interpreting coefficients jointly rather than individually
Acknowledging the uncertainty in individual coefficient estimates
Considering whether your research question truly requires separating the effects of correlated predictors

For interaction terms, high VIF values are expected and often unavoidable. This is sometimes called “inessential ill-conditioning” (Francoeur 2013). Centering the component variables can sometimes help reduce VIF values for interactions (Kim and Jung 2024).

Consider the substantive context: Sometimes, multicollinearity reflects important aspects of your data or research question. Removing variables to reduce VIF may actually harm your analysis by omitting important confounders or by changing the interpretation of remaining coefficients (Gregorich et al. 2021).

16512

Reposted by Magnus Johansson

Peter Tennant @pwgtennant.bsky.social · 28/05/2026

The methods section is the most important section of a paper. Without knowing the methods, the results are literally meaningless. So it's not OK to relegate the methods to the end. Because this suggests they are an optional afterthought rather than the essential core of the entire paper.

5507

Magnus Johansson @pgmj.bsky.social · 21/05/2026

Is it just me, or are Zenodo DOI links (for GitHub README markdown files) broken? I just get a broken link symbol, even for repos that used to work. If I use the image link separately, it renders fine in the browser. Example repo: github.com/pgmj/easyRas...

github.com

GitHub - pgmj/easyRasch2jmv: A Jamovi module of R package easyRasch2

A Jamovi module of R package easyRasch2. Contribute to pgmj/easyRasch2jmv development by creating an account on GitHub.

001

Magnus Johansson @pgmj.bsky.social · 19/05/2026

I think this is relevant for most complex tasks.

020

Reposted by Magnus Johansson

Francisco Rodriguez-Sanchez @frodsan.bsky.social · 17/05/2026

#StatsPubs No matter how many times you read it, this paper on hierarchical GAMs keeps giving. Clear model explanations together with R code (mgcv). So useful! doi.org/10.7717/peer... Thanks for writing @ericjpedersen.bsky.social, D. Miller, @gsimpson.bsky.social @noamross.net

In this paper, we discuss an extension to two popular approaches to modeling complex structures in ecological data: the generalized additive model (GAM) and the hierarchical model (HGLM). The hierarchical GAM (HGAM), allows modeling of nonlinear functional relationships between covariates and outcomes where the shape of the function itself varies between different grouping levels. We describe the theoretical connection between HGAMs, HGLMs, and GAMs, explain how to model different assumptions about the degree of intergroup variability in functional response, and show how HGAMs can be readily fitted using existing GAM software, the mgcv package in R. We also discuss computational and statistical issues with fitting these models, and demonstrate how to fit HGAMs on example data. All code and data used to generate this paper are available at: github.com/eric-pedersen/mixed-effect-gams.

216252