Rémi Thériault

What I don't understand is how does the Prior Scale value translates to standard deviation (SD) in the text? And right now `report_text.stanreg` is defined as `report_text.lm`. Do we need...

If going for the last option, it can be solved with a simple collapse by adding this on line 83: `values of 500 iterations and a warmup of 250) to...

Thanks for the inputs and explanations @DominiqueMakowski @mattansb! I'll get started on it shortly!

Ok so I: (1) changed the approach from formula interface to model-matrix; (2) added the `effectsize_table` class; (3) added a pretty printing method; (4) added info on dominance analysis; and...

I wonder if we should merge this PR without the confidence interval for now, and open a second PR to deal with the CI? Since it seems fully functional without...

After further investigation, it seems that the CIs are more reasonable when choosing "two.sided" or "less", rather than the default "greater". The lower bound is still always zero, but I...

When running `effectsize::F_to_eta2` (which makes use of the `alternative` parameter), the following message is printed when using `alternative = "greater"`: ```r - One-sided CIs: upper bound fixed at [1.00]. ```...

Awesome, great work! > Alright, I've adapted the non-F based CIs from Alf and Graf (1999), and expanded on the docs a bit. So now we have an actual CI...

> Hi @rempsyc if you haven't yet, I suggest reading the [One Sided CIs section in the CI docs](https://easystats.github.io/effectsize/reference/effectsize_CIs.html#one-sided-cis) - I think @bwiernik wise words will help clear up the...

> I think we can do without the section implicitly - we do link to it from the `alternative` argument: Ooooh I see I missed it because in the local...