ManyLabs replication variability

@RuxandraTeslo You should differentiate between Kahneman (research he did himself) and Kahneman (stuff he put in his books). His own research performed well in the big manylabs replication attempt, however you can't say the same for most of the priming stuff in Thinking Fast/Slow.

We were able to analyse 8 studies as pre-registered. Sometimes, we were able to replicate the original results but not the ManyLabs one, and vice versa. Overall, we replicated 37.5% of the original and the ManyLabs results.

They reveal that it was impossible to maintain allocation concealment in most of the centers involved in their manylabs-style replication, because the substance used to induce pain was yellowish, viscous, and induced obvious changes in the paws it was injected into

Scannies on their way to @icep2023 with the nightjet 🚆 Check out their talks on 22nd: @_maxsteininger (missing at this pic) at 11:20 on the effect of nature on pain perception @kim_doell at 15h on the Climate Manylabs Megastudy @boryana_td at 15h on pro-environmental behavior

‼️ New preprint on the psychology of AI (GPT3.5) is up (with @PtshPark & Zhu)! 🔴 We ran ManyLabs 2 studies with GPT3.5 🔴 Replication rate of 37.5% 🔴 We also find what we call the 'correct answer' answer in 6 studies, where we observe zero variance https://t.co/5JJnMupfOi

Talking about #ManyLabs projects at #SPSP2023 with the great @vlasceanu_mada @jgvoelkel

It was a pleasure to present our #ManyLabs project on trust in science at @Northeastern yesterday and to meet @smconstantino and her lab! 😊

There is OSC2015, ManyLabs 1-3, Camerer et al.'s Social Science Replication Project, @datacolada's Data Replicada, Soto's project on replicability in personality science, studies from @AaronCharlton's Marketing Replication Studies catalog, and many many more. (2/3)

Challenges like limited funding and infrastructure affect African researchers. But initiatives like ManyLabs Africa show what’s possible. Explore these stories in A Better How. Support research that breaks barriers. Get your physical copy at Nuria Bookshop in Nairobi or order

The ManyLabs Africa initiative faced numerous challenges—limited infrastructure, high research costs, and low representation in global studies. Yet, it also uncovered practical solutions to these barriers. From improving access to open science practices to fostering equitable

日本のデータコレクションを手伝ったManylabs Climate Projectの研究が、Science Advances（@ScienceAdvances）にアクセプトされました。 / Vlasceanu, Doell... Van Bavel. (accepted). Addressing Climate Change with Behavioral Science: A Global Intervention Tournament in 63 Countries.

I’ve already made use of the AIID/Attitudes 2.0 dataset. The Ideology 2.0 dataset isn’t suitable. I know of https://t.co/xnqmLj7vEf, the ManyLabs and psych science RRRs.

Work spearheaded by @BrianNosek + collaborating psych labs crucial for bringing this crisis into full view The ManyLabs project proved @DataColada's methodological critiques seriously undercut most TED talkable studies The term Repligate was born https://t.co/sV3Z6g7AJz

It is our great pleasure to announce our next @CBehav Seminar, where we get to enjoy a talk from @kim_doell, a newcomer to our community. Join her talk "The Manylabs Climate Collaboration" on Monday 17 June 17th at 12 in ZT702 @UniKonstanz. https://t.co/nxxO8qPAyr

How do people perceive #science and its role in society? Do they #trust scientists? 🧑🏾‍🔬 We did a global 67-country survey (n=71,417) to test this, working with 239 researchers from all continents – the #TISP #ManyLabs study. It’s now available as #preprint: https://t.co/qFNmD9IeY4

Great to be back @oiioxford for some days – and visit @LSEnews: In @LSE_PBS‘s Public Understanding of Science seminar, I gave a talk on our #ManyLabs study #TISP and presented results on populism and science communication 🌏 Thanks for having me, Martin Bauer & @melaniesmallman!

manylabs project to replicate taylor swift lyrics when

A little over a year after our call for collaboration, we just submitted the main article of our global #ManyLabs study on #trust in science 🥳🥳 237 co-authors keep your fingers crossed 🤞 @nielsmede https://t.co/dChlb6RNuJ