So many questions! So many thoughts! Would love to discuss although I am no expert.

I think for submissions you can upload a pdf and once you're accepted I think it's not hard to get rtf or html (which imports to word) from latex.

Overleaf works pretty well, I think even for non-experts, once you set up the document it's not hard to collaborate productively.

I knew the basic idea, but the article does a good job of bringing the message home. For anyone who feels like their world is collapsing: a teach a course on Bayesian statistics!

This is great! Thanks! Now I just need one for people who want to keep fighting over Bayesian vs frequentist. We're a small group and it's hard to find each other.

Thanks for the tip. From website it's hard to figure out anything concrete. There's a lot beauracratese and buzzwords. Is there any actual progress? coara.eu

The journal system (and peer review) are still university and granting agencies go-to evaluation tool. Are there good ideas around on workable alternatives?

That's news! I thought they'd found it already!

Depending on what you mean by "reasonable likelihood", this is either untrue in general or meaningless.

doi.org/10.1177/2515...

Good question! Power and significance relate to testing. Estimate of effect size doesn't: CI of effect size should be valid at small sample size, but don't use it for testing. In any case: for prospective power analyses use smallest effect size of interest and don't use an estimate.

The appropriate effect size to use is the smallest you care about. You cannot take seriously significant results for which you are underpowered so choosing sample size also chooses smallest detectable effect.

Isn't it more accurate to say: "our evidence for an effect failed to reach significance"?