2/2 Thank you, Patrick Butlin, and the entire group for this journey 🙏

Sorry - I missed this comment! And indeed, we are talking about all comparisons that are meaningful, which is defined by the questions. We recommend constructing the prediction tree a priori. Sorry if that wasn’t clear

Thank you Johannes 🙏

3/3 See the preprint here, using the data from @briannosek.bsky.social’s reproducibility project: arxiv.org/pdf/2006.11585

2/3 and suggest a method - treeBH - that allows one to do so without losing power and while taking into account the hierarchical structure of the tested hypotheses. In another paper that is still under review, we show that applying this method might reduce the number of non-replicable findings. >>

Super interesting! Congrats 🎉

It’s never too late :-))

Over 2.5 million Israelis joined today’s anti-Netanyahu protests, organisers estimate. 300,000 are now in Tel Aviv's Hostage Square. (Haaretz)

sounds like a good deal :-))