Nearly as bad as Americans naming their universities, such as Washington University (Missouri, no Washington State or D.C.) and Miami University (Ohio)

Agree, this paper has a series of figures that lead to a 3d representation of 3 axes I work out, so this builds up logically and it is not about interpreting the purely example dataset, but getting the idea of clustering in 3dimensional space. So I think the figures should be fine here.

Figures go on OSF always anyway to make sure I have them licensed (CCby4) before journals get their dirty hands on licensing them ;). Turns out the earlier license counts ^^

hahaha

"The implication is that some hypothesized or preregistered effect sizes may be impossible. At the same time, these restrictions offer a way of statistically triangulating the plausible range of unknown effect sizes."

Yes that's pretty much exactly it I think, thanks

(or more likely, I'm just missing smth, super jetlagged here haha)

In the 2023 Royal Society paper, how do you define 'reproducibility'? eg "The true reproducibility rate of a result is a function of not only the true model generating the data but also the elements of the idealized experiment" make little sense given my definition so we must have different ones

I've seen one of these two and not connected this but I think you're right, this is exactly on point! Thanks a lot, Berna

I've seen arguments made that e.g. a statistical effect or model has "low replicability", but if you actually sit down and calculate what the replication ceiling is given constraints, the results can be pretty close to the theoretical maximum.

Are you aware of related concepts? Thanks!

🔥

Thanks Rob, that's helpful!

Oh Paul that's just silly

This is a projection of 3 principal components (the 3 axes) into a 3dimensional space. I think this will be tricky in 2d given that I have ... 3axes.

Publishing a paper there costs 0, of course. The open access fees are steep, but researchers at Dutch unis are protected under the taverne amendment, which is why you can find the full text on my website.

Folks have cited the DSM-5 field trials paper by Regier et al 2013 paper quite a bit, but I always felt it isn't quite appreciated how ridiculously low kappa=0.28 actually is, hence the figure.

Here a paper version rant on this in Nature Reviews Psychology.

www.nature.com/articles/s44...

(PDF: eiko-fried.com/wp-content/u...)

Thanks for the context Casper