Same five studies. Same treatment effect. One meta-analysis (raw): I² = 0%. Another meta-analysis (SMD): I² = 94%. How can both be right? 🤔 Find out in my latest blog post👇 mzloteanu.substack.com/p/standard-n... #Statistics #MetaAnalysis

We've got two online IPD meta-analysis training courses this Autumn, with places still available - one is a general introduction about doing IPDMA projects (the whole process, not just the stats bit) & the other is about the statistical modelling aspects - more here: www.ipdma.co.uk/training-cou...

Training courses | IPD Meta-Analysis

ipdma.co.uk

Every finding rests on a stack of choices: which scale, which model, which studies to pool. Many are arbitrary yet reasonable, and they can change the conclusion. I started a series with essays walking the research pipeline one step at a time, plus a living collection of ~90 multiverse papers.

New series: Arbitrary yet Reasonable

I am starting a series of essays called Arbitrary yet Reasonable. It is about the many defensible decisions behind every scientific finding: which instruments, which statistical models, and which vari...

constantinyvesplessen.substack.com

Announcing roxygen2 8.0.0 for #RStats package developers! 📦 This major release streamlines documentation workflows and includes over 100 bug fixes, S7 support, R6 tooling, a lighter footprint, and modern configuration. Learn more: opensource.posit.co/blog/2026-05...

roxygen2 8.0.0

roxygen2 8.0.0 is now on CRAN, with first-class support for S7, a bunch of R6 improvements, a new configuration style, and a slimmer dependency footprint.

opensource.posit.co

The #OpenScience working group* of the German Psychological Society's section on #ClinicalPsychology and Psychotherapy has published a paper on legal challenges and solutions for #OpenData, German #PrePrint here (paper just accepted): psycharchives.org/en/item/7c9d... #GDPR #PsychotherapyResearch

We've written about AI-induced physician deskilling that has already surfaced thelancet.com/journals/lan... @tberzin.bsky.social AI-induced never-skilling among newly trained doctors, while not yet proven, is a serious concern that needs to be addressed nature.com/articles/s41...

I spent the past semester building interactive teaching apps with AI-assisted coding. I used my personal website to share them with my students, but that makes them hard to find for other educators. So I built Galvoro — galvoro.app

Wimsatt: Perhaps, just perhaps, the best explanation of a phenomenon at a given level is to be found at the same level. Kendler: Noone would seek to explain hypertension at the level of quarks. A few of my favourite papers on this topic of absurd biological reductionism: 🧵

Brent W. Roberts@bwroberts.bsky.social · 6mo ago

This...the fetishizing of unreplicable neuroscience in order to legitimize our work is self-defeating. We don't need to know that depression is in the brain (as opposed to what???) to know it is problematic. How much money have we wasted proving something like depression is "real"?

A major release of the RoBMA package (4.0.0) is now on CRAN. Main highlights: - fully redesigned interface closely matching metafor - dedicated brma, bselmodel, bPET, bPEESE for single model fits - fit diagnostics features (residuals, influence, etc) (overview at fbartos.github.io/RoBMA/articl...)

@taymalsalti.bsky.social has done trojan work trying to reproduce the results of three meta-analyses, finding the usual host of issues of effect sizes that are apparently incorrect or non-reproducible Our article is out now: doi.org/10.1111/ejn....

Results of this survey will be scientifically uninterpretable: it contains leading questions and is advertised using leading imagery. Disappointing to see such a nice opportunity wasted to study this interesting and important topic.

British Psychological Society@bpsofficial.bsky.social · 6mo ago

What should the minimum age be for social media access? This is one of the key questions in the government’s consultation on children’s online experiences, including how we support safer & healthier digital environments. Help shape our response: forms.cloud.microsoft/pages/respon...

The Danish FKU is a ‘psychometric test’ to assess parental competence and has resulted in 6% of Greenlandic children in Denmark being removed from their families. It contains among others knowledge and intelligence items and the Rohrschach (!) test. en.wikipedia.org/wiki/For%C3%...

🧵 New WP with @kaimiele.bsky.social. We document 2 unsettling patterns in mental health care: 1) despite universal coverage, few individuals with mental illness receive guideline-consistent treatment, and 2) the more severe the illness, the lower the treatment uptake & the longer the wait times. 1/n

Want to feel sad and hopeful at once? What´s needed beyond Open Science? "All of the ingredients for low quality occurring at some non-negligible rate seem to already be in place"—the nicest way I've ever heard someone say a terrible true thing. What a great pitch for forensic meta-science ✌️

ReproducibiliTea@reproducibilitea.org · 10mo ago

📽️ "Taboo topics we can no longer avoid: negligence, tampering, and fraud" / with Ian Hussey @ianhussey.mmmdata.io youtu.be/sdRRafephyM?... The event was organised by ReproducibiliTea UniBasel @swissrn.bsky.social

"The rewards to 'discovering' a spectacular scientific finding [in psychology] are large; the rewards to debunking frauds or deflating exaggerated claims are small if not non-existent. If these are the rules of the game, we should not be surprised at the way the game is played."

When psychologists mislead us

From Piltdown Man to the Stanford prison experiment, many famous scientific discoveries have been exposed as hoaxes or distortions

ft.com

Many innovations of psychotherapies are currently examined. See our overview in the last issue of World Psychiatry: (1) Digital interventions (2) Personalized treatments (3) New/ improved therapies (4) Dissemination With @mathiasharrer.bsky.social @toshi-frkw.bsky.social doi.org/10.1002/wps....

Innovations to improve outcomes and uptake of psychotherapies for mental disorders: a state‐of‐the‐art review

Psychotherapies have been found effective in the treatment of most mental disorders. However, substantial improvements are still much needed, and many innovations of therapies are currently being dev...

doi.org

📽️ "Taboo topics we can no longer avoid: negligence, tampering, and fraud" / with Ian Hussey @ianhussey.mmmdata.io youtu.be/sdRRafephyM?... The event was organised by ReproducibiliTea UniBasel @swissrn.bsky.social

Open Science Basel Year 6 #4 - Ian Hussey

YouTube video by BAMM

youtu.be

New preprint 📄 How many psychotherapists does Austria need? This one is special to me beyond the topic: it’s an update of the very first paper I ever wrote, almost a decade ago. And I’ve always wanted to open a paper with a joke — this time I finally did. Let’s see if it survives peer review.

After 5 years of data collection, our WARN-D machine learning competition to forecast depression onset is now LIVE! We hope many of you will participate—we have incredibly rich data. If you share a single thing of my lab this year, please make it this competition. eiko-fried.com/warn-d-machi...

WARN-D machine learning competition is live » Eiko Fried

If you share one single thing of our team in 2026—on social media or per email with your colleagues—please let it be this machine learning competition. It was half a decade of work to get here, especi...

eiko-fried.com

Fifteen of those twenty years were spent writing unsuccessful grant applications.

Abby@arobertswrites.com · 11mo ago

You can tell that LotR was written by an academic because Gandalf disappears for like 20 years doing research to answer a single question.

"(...) observed changes in PHQ-9 scores over treatment weeks may, at least partly, reflect shifts in how participants engage with the scale rather than true changes in depressive symptomatology." link.springer.com/article/10.1...

Why are we still using the PHQ-9? A Historical Review and Psychometric Evaluation of Measurement Invariance - Psychiatric Quarterly

The Patient Health Questionnaire-9 (PHQ-9) is a widely used tool for assessing depressive symptom severity and as a screening tool in the diagnosis of major depression. Designed as both a diagnostic i...

link.springer.com

What are some articles, blogposts, essays, etc. that you still think about over a year later? Here are a few of mine:

Yes, I am one of those horrible people who submit during the season. New preprint, 1st paper from my #ERCStG DECOMPOSE www.medrxiv.org/content/10.1... The goals of DECOMPOSE are break-down psychological interventions into components, integrate these in a taxonomy and evaluate incremental efficacy

Accessibility and public availability of treatment protocols and manuals for evidence-based psychological interventions for severe mental disorders: A survey of randomized trials

Background: Psychological interventions are developed and tested as packages of multiple components. Identifying active ingredients requires access to detailed intervention descriptions, such as from ...

medrxiv.org

I am seeing this paper cited everywhere, @sciencenews.bsky.social @naturenews.bsky.social conferences, soon policy papers osf.io/5rf2m/ First, what is the method of this paper? Systematic review? Scoping review? Overview of reviews? Narrative review? Or none was adequate?

HOW MUCH SCIENCE IS FAKE?

Hosted on the Open Science Framework

osf.io

JBI, Cochrane and Campbell Collaboration are partnering to co-facilitate an initiative, Building a Global Evidence Synthesis Community. We are calling for expressions of interest to join us as we seize a once-in-a-generation opportunity for a global step change in #EvidenceSynthesis synthesis. 🧵1/

Stigmatising and inaccurate portrayals of mental disorders: Researchers asked generative AI models to depict different disorders. Images reflected historical biases & pop-culture references rather than evidence-based clinical representation of these conditions mentalhealth.bmj.com/content/27/1...

Starting my job search: I just finished my PhD in Clinical Psychology and am looking for opportunities to make an impact. I have expertise in meta-analyses, advanced stats, psychometrics, R, and project management. If you know of a position that might be a great fit, please feel free to reach out.

Do a DAG Before You Do a Specification Curve Analysis #CausalSky #Stats replicationnetwork.com/2024/11/18/r... With an example of reducing >1,200 specifications to 6. Computational power should not replace theoretical thinking and model building. 1/

REED: Do a DAG Before You Do a Specification Curve Analysis

My previous two blogs focused on how to do a specification curve analysis (SCA) using the R package “specr” (here) and the Stata program “speccurve” (here). Both blogs provided line-by-line explana…

replicationnetwork.com

TIL blue zones of longevity are not real: “Newman’s main argument is that a significant proportion of supposed centenarians may simply not exist. His research shows that around 1900, when the U.S. started to issue birth certificates, the number of centenarians aged 110 or older dropped sharply…”

Eric Topol@erictopol.bsky.social · 2y ago

There's really no such things as "blue zones" for longevity www.science.org/content/arti... open-access