HPI, mythes et faits autour de l‘intelligence clarifiées par deux neuropsychologues, Sophie Brasseur et Catherine Cuche! ecoutez ce nouvel épisode de @milgrampodcast.bsky.social, avec une nouvelle identité sonore! open.spotify.com/episode/20RI...
To me, the main strength of frequentist is to manage type I and type II errors. Also for simple model the computing efficiency. Does anyone know if type I/II can be adressed using bayesian?
Pretty proud of this one — it represents 3 years of work!
Our new publication in American Psychologist uses EEG to investigate whether the brains of rescuers respond differently than former perpetrators and bystanders.
@erc.europa.eu @ghentccn.bsky.social
article: psycnet.apa.org/fulltext/202...
New paper, on a worrying trend in meta-science: the practice of anonymising datasets on, e.g., published articles. We argue that this is at odds with norms established in research synthesis, explore arguments for anonymisation, provide counterpoints, and demonstrate implications and epistemic costs.
Thanks, glad it's of interest to you. Below are some of the references. If you have further questions, do not hesitate to reach out
[1] www.frontiersin.org/journals/psy...
[2]
link.springer.com/article/10.3...
[3]
rips-irsp.com/articles/10....
Wahou, so refreshing this first day at the meta-science conference #PSE8. How did I change my mind from science is broken to yes but great minds are working on it!
I also received very positive feedback on my work on outlier detection (that should be out as a preprint soonish)
Now Aurélien Allard with the second talk of #PSE8: How is it possible that there is massive scientific progress, even thought there is a lot of scientific error?
The Impact of Non-Neural Sources on Aperiodic EEG Activity
www.biorxiv.org/content/10.6...
We're running a 5th edition of the always-exciting UCL Summer School on Consciousness and Metacognition this year, 8th-10th July 2026 in London. Accommodation and travel expenses are covered.
For more information and how to apply, check out metacoglab.org/summer-schoo...
A hypothesis developed based on the data is often more likely to be true, than if you had not used the data.
The problem is not whether the hypothesis is true.
The problem is the hypothesis was not severely tested. You can't *claim* it is true until you test it on new data.
main goal for this year: find a new job! 🙂
looking for a role with fun & complex technical challenges & within a great community. my main expertise is in signal processing/EEG/MEG, but topic-wise I am quite flexible.
science/industry both great! starting mid-year. nschawor.github.io/cv
Interesting, and good recommendations that apply broadly « Increase methodological rigour: increase sample sizes; pre-register analyses; use transparent, harmonised pipelines; validate models externally; share data and code; and publish null results.» www.cambridge.org/core/journal...
It seems massive but I am not a fMRI expert. They seems to report that the relation between bold signal and neural engagement is not straightforward, and can even be reversed. So increase bold = increase vascular activity, but not necessarily increase neural activity www.nature.com/articles/s41...
course schedule as a table. Available at the link in the post.
I'm teaching Statistical Rethinking again starting Jan 2026. This time with live lectures, divided into Beginner and Experienced sections. Will be a lot more work for me, but I hope much better for students.
I will record lectures & all will be found at this link: github.com/rmcelreath/s...
The decline of democracy in the US means you can’t visit the US if you commented on the decline of democracy in the US
But completely agree with the increase of very complex model (GAM - ML) which ask the same question from one study another but use different models, making empirical contributions statistically difficult (impossible ?)
Ah I misunderstood. For me, the main argument is power at first place. If I have a 3x3 within design, with unbalanced observation due to absence of observation or else, mixed models are the right things to do to keep all information. However, a 3x3 random structure is very difficult to model.
In another words, I am wondering if we can tease apart bayesian estimation / maximum likelihood estimation, from bayesian / frequentist inferences. While Bayesian estimation seems to be more powerful than MLE, frequentist inference seems also more powerful (as they can do all of bayesian inf + more)
For meta-analysis, I never saw one, but I just found that pmc.ncbi.nlm.nih.gov/articles/PMC....
My initial question is about type 1 et type 2 error, which are useful frequentist construct. However, can they be used if the estimate are drawn with bayesian estimation? If not is this developped?
Can frequentism inference operate under bayesian estimation? With the development of mixed (generalized) linear models, bayesian estimation show to be more able to fit more complex model (e.g, non normally distributed phenomenon). Once we have this estimates, is frequentism inferen. still an option?
new paper in TICS officially out today. great learning from and writing with Anastasia, and super cool cover art from Prof. Pinar Yoldas.
www.cell.com/trends/cogni...
⛓️🔒 La prison semble être une réalité lointaine… Mais les délinquant·es y sont condamné·es en notre nom à tous·tes. Dans notre nouvel épisode 100g, @elodiekox.bsky.social explique les conséquences psychologiques et cognitives sur les déténu·es. 🧠
👉https://milgram.ulb.be/100g/episodes-100g/prison/
Great piece on the absurdity of brute force multiverse analyses.
www.pnas.org/doi/10.1073/...
We used recent tools to analyze the prevalence of the effect (Bayesian Prevalence elifesciences.org/articles/62461 developed by @robince.bsky.social) within our sample, and observed that some participants implictly associate a positive valence with agency, while others associate a negative one.
Our study suggest that we explicitly say yes, but implicitly we seem not to attribute a specific valence, or at least, it depends on who is experiencing agency.
New preprint! Do you prefer to act rather than not to act? We investigate whether the sense of agency is valenced, with Inès Mentec; Charlotte Auger, Murielle Coulibaly, Martine Malaise, Matéo Vandeville, and @axc.bsky.social leveraging explicit and implicit methods.
osf.io/preprints/ps...
New research in Rwanda has started ! We continue to try to understand what differs those who participate in a génocide, and those who instead risk their lives to save others.
Here is a picture of our novel and ecological risk-taking task, inspired by the BART.
How does the brain decide? 🧠
Our new @nature.com paper shows that neural activity switches from an 'evidence gathering' to a 'commitment' state at a precise moment we call nTc.
After nTc, new evidence is ignored, revealing a neural marker for the instant when the mind is made up.
rdcu.be/eGUrv
Happy to share a new preprint in which @paulbuerkner.com and I introduce a novel model-based approach for precisely estimating the onset and offset of M/EEG effects!
www.biorxiv.org/content/10.1...
