Model parameters about easystats HOT 4 CLOSED

omega squared etc.

I hate these ;-) I'm no ANOVA fan and have no idea why I should use it, and the functions around anovas in sjstats (like eta_squared() etc.) are really stepmotherly treated by me ;-)

from easystats.

This is very close to the idea of the broom-packages. We have to think about the benefit / surplus this package would have compared to broom / broom.mixed, so the package will be recognized.

from easystats.

Yes. For report, (right before the creation of easystats), I actually thought of using broom as a basis, to which I added other things.

I think that the main benefit of this package/part of a package would be providing this other stuff, which could include:

• effects existence indices: p values if not there originally, but ropes and others for Bayesian
• Different CIs (bootstrapped...)
• Effect sizes indices (standardized coefficients, omega squared etc.)

from easystats.

Haha I totally agree... the only reason why I am thinking about ANOVAs is due to their (still existing) widespread use in psychology. So I tell myself that neophytes will come and find out their ANOVAs can be easily done with R, and I hope to further encourage them using better procedures (like here when I tried to trick newcomers into using mixed models 😈). My psycho package gained quite some users after this post.

So yes if we implement these it we should put a minimal amount of effort in it. And it shouldn't be a focus. But it could help with the popularization in psych universities (for oldschool researchers and teachers/students (in Paris they still teach students how to do ANOVAs by hand... 😢))

from easystats.