We are thrilled to announce the Paul Meehl Graduate School for metascience at Eindhoven University of Technology. We will offer 6-8 free workshops a year for metascience PhD students. You can register and find more information at paulmeehlschool.github.io

RT @maltoesermalte
Ok but it seems like an omission of important contextual information that his mother's face looks like this: twitter.com/Variety/status/165

RT @niclas_kuper
My dissertation dealt with the topic of person x situation interactions across different study designs and timescales. For the three manuscripts, see the linked threads below!

This guy is often shown as the illustration of the Dunning-Kruger effect. Supposedly MacArthur Wheeler, the 'invisible' bank robber. But he doesn't fit the wanted ad at all. Is he a different incompetent bank robber? I guess there are no pictures of MacArthur Wheeler?

RT @salonium
Me, @StuartJRitchie, @Evolving_Moloch and @rubenarslan having nice Greek food last week and talking about the replication crisis and Twitter drama.

RT @LeonidTiokhin
⭕ Red Teams ⭕ are undervalued, but are vital for organizations dealing with adversaries and dynamic environments. This is why Red Teams were recently used by OpenAI in their development of GPT-4.

[ THREAD ]

Oh no he found our holdout data
---
RT @bmdv
Ungenutzte nützen nichts. Der 🇪🇺 hilft uns, das zu ändern. Mit ihm sollen maschinengenerierte Daten leichter verfügbar und fair genutzt werden können ➡️ für den ​en Fortschritt und eine klimafreundliche Zukunft. @Wissing
twitter.com/bmdv/status/165194

RT @Leesplez
If there's one brain disease I swear all psychologists have, it's one that causes them to spend years of their life developing their own scale, then publishing it without an accessible, final version anywhere in the manuscript or online.

RT @ianhussey
There is a worrying trend in meta-science where authors are unwilling or unable to state which articles are subject to which critiques, as if only the prevalence of issues across articles is relevant and not the specific instances of these issues.

Scrutiny is the goal! twitter.com/sharoz/status/1650

See also "No evidence that economic inequality moderates the effect of income on generosity" by @StefanSchmukle
pnas.org/doi/abs/10.1073/pnas.
and this follow-up where a dose of preregistration cures the p hacking pnas.org/doi/abs/10.1073/pnas.
---
RT @StuartJRitchie
New from me: remember that super-famous psychology study from 2012 that claimed that higher social-class people were more u…
twitter.com/StuartJRitchie/sta

RT @ianhussey
I recall reading a paper, probably before the pandemic, about the myriad ways in which SES is measured, that there was little content overlap or convergent evidence between them, and yet they were treated as equivalent measures of "SES". Does anyone recall the paper?

RT @EikoFried
Join our team as a PhD student :).
I'll be co-supervisor for this project, together with the supervisor team Anne Roefs & Lotte Lemmens located at Maastricht University. twitter.com/EatingLab/status/1

RT @paulbuerkner
I am hiring a PhD student or Postdoc to work with me and my group on Bayesian workflow topics, this time specifically on simulation-based Bayesian inference. Check out paul-buerkner.github.io/positi for more information. If you have any questions, please just respond to this tweet.

RT @WDRaktuell
Sexuelle Belästigung, Demütigung oder Mobbing - dafür ist offenbar auch der Wissenschaftsbetrieb besonders anfällig. Ein Grund, so Historikerin @cgpublichistory: das "sehr starke Abhängigkeitsverhältnis" zwischen Profs und Mitarbeitenden.
www1.wdr.de/nachrichten/landes

One reason why I took my wife’s name..
---
RT @henninghermes
Our findings reveal that migrants receive 4.4 percentage points fewer responses. Moreover, their responses contain fewer slot offers, are shorter, and are less encouraging. This discrimination likely perpetuates existing inequalities for disadvantaged children. 3/9
twitter.com/henninghermes/stat

RT @logopetria
@CarcassiFausto - Difficult to solve the problem of alignment

RT @CarcassiFausto
What if LaTeX was the real AI risk all along?

- Completely obscure, uninterpretable results
- Manipulates its users into enthusiastically propagating it
- Probably an x-risk

Show older
Mastodon

a Schelling point for those who seek one