Want to start off your Shiny app with a modern, sleek look? ✨ Use the {bslib} package for your UI: straightforward card system, Bootstrap 3/4/5 support, and tons of built-in components. 📦 rstudio.github.io/bslib/index.... #RStats

Some researchers don't discuss their future research plans for fear of being scooped. Not me. I drop bad ideas for unscrupulous people to 'steal'. - What are the neural correlates of Open Science practices? - What is the role of habits in learning a new skill through repetitive practice?

Science writers, as the white paper elucidates, use an article structure called the "5 Bits" (screenshot 1). The prompts given to the LLM (screenshot 2) do not specify this. They do not provide good examples (as in one- or few-shot prompting). They generally do not follow prompting best-practices.

Many of the responses to this paper are a version of "Obviously, this cannot work!!" -- I find this a baffling response, to be honest. I don't think it is particularly "obvious", unless you're very anti-curious, but more importantly, I don't know what "this" is.

So what’s the take-home? 👉 Silicon samples are highly sensitive to analytic decisions. 👉 There’s no single “best” workflow. 👉 Different defensible choices can lead to wildly different outcomes. And to date, we don’t have much work to help motivate any choices.

All of the silicon sample configurations were only, at best, moderately correlated with the human data when it came to preserving the ranking of participants. And many of them were negatively correlated with the human data.

Can large language models stand in for human participants? Many social scientists seem to think so, and are already using "silicon samples" in research. One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want. THREAD 🧵

Very happy to see this out. Use of LLMs to simulate study participants is something we'll see increasingly, if we want it or not. This paper is a super helpful guide to understand the impact of seemingly minor or unimportant decisions while doing so!

I'm afraid we've been at it again

Whoa—my book is up for pre-order! 𝐌𝐨𝐝𝐞𝐥 𝐭𝐨 𝐌𝐞𝐚𝐧𝐢𝐧𝐠: 𝐇𝐨𝐰 𝐭𝐨 𝐈𝐧𝐭𝐞𝐫𝐩𝐫𝐞𝐭 𝐒𝐭𝐚𝐭 & 𝐌𝐋 𝐌𝐨𝐝𝐞𝐥𝐬 𝐢𝐧 #Rstats 𝐚𝐧𝐝 #PyData The book presents an ultra-simple and powerful workflow to make sense of ± any model you fit The web version will stay free forever and my proceeds go to charity. tinyurl.com/4fk56fc8

You're very welcome @vincentab.bsky.social P.S. Check out our preprint on an alternative to staring at coefficients: j-rohrer.github.io/marginal-psy...

Waiting for my preprint to be accepted, so in the meantime a teaser: here's what happens when you try to estimate a between-scale correlation based on LLM-generated datasets of participants, while varying 4 different analytic decisions (blue is the true correlation from human data):

Proposal for how to fix family wise error rates. For every uncorrected p value you must add an extra letter to the claim. “Eating chocolate maaaaaaaaay be associated with lower rates of stroke”

Great thread. At some point, we have to collectively decide: do the numbers matter or not? So many claims are made on the basis of The Numbers, and yet when the numbers are are wrong or impossible the claims rarely change. So what purpose do they serve in the first place?

This is textbook p-hacking! I'm not making any assertions about the intent of the authors. They state in the text that this was an ad hoc adjustment to the specification based on looking at the results. There's no other way to describe this.

New paper out with @boryslaw.bsky.social 🥳 In which we sketch out how to rethink measurement invariance causally for applied researchers. And provide a causal definition of measurement invariance! www.sciencedirect.com/science/arti...

This is schmeasurement's Dead Salmon fMRI study: A cautionary tale of how standard research practices can produce misleading results. I love it. Immediately putting it on my reading lists.

In a new paper, my colleagues and I set out to demonstrate how method biases can create spurious findings in relationship science, by using a seemingly meaningless scale (e.g., "My relationship has very good Saturn") to predict relationship outcomes. journals.sagepub.com/doi/10.1177/...

2019 national study found ~150 U.S. minors had gender-affirming surgeries via insurance claims. 97% were cis boys getting breast reductions for gynecomastia, not trans youth. No surgeries were done on kids under 12; procedures on ages 13–17 were nearly all chest-related. #TransYouth #JAMA

Nearly 3 months since I contacted both the authors and editor of @jama.com about failure to share data. Article states data and code is available, but the github repo is empty. Not private or partial, empty. Has never contained anything. No replies from anyone. jamanetwork.com/journals/jam...

For Monday crowd. INSPECT-SR, a tool to help you identify problematic (inc. fraudulent) RCTs, now available.

www.medrxiv.org/content/10.1... INSPECT-SR: A tool for assessing trustworthiness of randomised controlled trials.

Nice article arguing that psychology must continue to learn from its past mistakes or we will create a new AI-flavoured replication crisis.

Today in "did we take this joke too far?": @malte.the100.ci @grinschglsandra.bsky.social and I will jointly teach two lecture series this year: Psychology of Digitalisation: Everything is Computer Digitalisation of Psychology: Computer is Everything (really) Also, new lab logo

Yet another example of original authors being allowed to say anything they want in replies to critiques unconstrained by verifiable facts. Reply states we didn’t consider things we explicitly did, and says the original article never said things it explicitly did. Read Jamie’s thread for details:

This has now been published in PNAS! www.pnas.org/doi/10.1073/... The original authors also posted a reply: www.pnas.org/doi/10.1073/... Quick thread on some additional thoughts and then I'm probably done talking about this one 🧵

Key claim of trimodal response to SSRI in the STAR*D trial is not computationally reproducible www.sciencedirect.com/science/arti...

This RedNote post shocked me today. An anonymous user says they are going to upload 60 papers to arXiv soon and are looking for “citation partners” to boost citation counts together by citing each other. Something is seriously wrong in the system of evaluating researchers by citation counts.

If you have R students learning to clean data, the truffle package by Ian Hussey ( @ianhussey.mmmdata.io ) might be for you. Plus the icon is a cute pig 🐷 Thanks, Nur Shahir, for the tip! www.linkedin.com/in/nmshahir/ github.com/ianhussey/tr... #rladies

Calls for Stronger Theory in psych should take note:

What if we thought about digital behaviour not in terms of the form that it takes but the psychological functions that it serves? In this preprint we introduce technological behaviourism, a theoretical approach grounded in operant conditioning. @shuhbillskee.bsky.social

A new lens for understanding how digital technologies facilitate human behaviour: https://osf.io/ztev4

Did you know there's a package to reformat code to meet the {tidyverse} style guide? {styler} is super useful for teaching students and helping them make their code more legible. It even has a drop down menu to do it with a point-and-click. styler.r-lib.org

I often find my students' code is littered with dependencies that are loaded and then not used. This is a bad coding practice, but it's also very boring to check. I updated {psychdsish} to include a function that does this for you.

New and improved 😆 Those who were at the first training will know we needed to 🤦‍♀️ See you tomorrow, we're so thrilled to see all of the interest in supporting @psyarxivbot.bsky.social

Listen, I'm gonna multiply these two coefficients. If you claim I'm doing a mediation analysis, that's on you.

Due to a recent influx of problematic submissions, PsyArXiv has switched to pre-moderating its content. If your submitted preprint had not yet been approved, it will be temporarily inaccessible to the public (you can still view your preprint when logged into your OSF account). #PsychSciSky

RDM Weekly Issue 9 is out! 📫 - NIH DMS Plans Lesson Set - Ethically Making and Sharing Data @cynthiaheider.com - Creating Psych-DS #rstats Projects @ianhussey.mmmdata.io - Yoda Metadata Editor @utrechtuniversity.bsky.social and more! rdmweekly.substack.com/p/rdm-weekly...

RETRACTED: Clinical and metabolic response to probiotic administration in people with Parkinson's disease: A randomized, double- blind, placebo- controlled trial It happens too many times, takes time and undermines science in general. www.sciencedirect.co...

{truffle} is an R package for teaching users to process data. Semi-realistic psychological datasets with predetermined effects (via `truffles_` functions) are then hidden in common data processing headaches (via `dirt_` functions) for students to clean and analyze. mmmdata.io/posts/2025/0...

OpenAI recently released GPT-5, and its smaller derivatives, GPT-5-mini and GPT-5-nano. On paper, GPT-5-nano is much cheaper than GPT-5-mini: $0.40 per million output tokens for nano vs. $2 per million for mini. But behind the scenes, nano is secretly costing the user almost as much as mini. 🧵

3. I then found out that the author mixed suicide RATES (per 100,000) with suicide RATE RATIOS either standardized to 1 or to 100. See my comments to the figure. I confirmed this by looking at several individual studies. Of course this doesn't make sense at all.

My general orientation toward FAIR-ification is to celebrate all incremental moves, so I'm fully in support of 'psych-DS...ish' tools while we're in the process of building more. And if anyone wants to get involved we are doing some more R-based work on the standard as we speak! github.com/psych-ds

PsyArXiv is seeking new moderators to help combat an increase in AI submissions! If you've ever posted a preprint to PsyArXiv, please consider joining. Minimum commitment 1h/month, there's a training session this Monday @ 1pm ET. More info here: forms.gle/9LB1rEtxHAeZ... #PsychSciSky

“Whoever read this paper, editors and the peer reviewers, obviously made no attempt to understand what was going on.”

I just heard about @renebekkers.mastodon.social.ap.brid.gy's Bad Science Bingo Card - quite a fun approach to a serious topic. detectingbadscience.wordpress.com/bad-science-...

I wrote an R package that creates standardized R project structures that are compliant with @mekline.bsky.social's psych-DS...ish. It also creates additional features for reproducibility and teaching like a readme, license, .gitignore and Quarto templates + can validate existing projects

👇 MetaROR is very proud and excited to announce this new partnership with @plosbiology.org 🙏 As part of this, "PLOS Biology will formally consider meta-research articles that are peer-reviewed on the MetaROR platform", in order to "improve transparency" and "[prevent] futile cycles of peer review".