Open jmgirard opened 11 months ago
In the long run, the goal is to use marginaleffects as backend (see #214), which will probably go along with adding more features.
For now, and if you're familiar with the syntax of ggeffects, you could also use the hypothesis_test()
function. See a series of vignettes and examples starting here:
https://strengejacke.github.io/ggeffects/articles/introduction_comparisons_1.html
Let's say I fit a linear model with a continuous-by-categorical interaction.
I can use
estimate_slopes()
to see what the slope of the continuous variable is for each level of the categorical variable.But I don't think there is currently a way in {modelbased} to contrast those slopes. However, you can do so with
emtrends()
.A {modelbased} wrapper for this functionality would be nice, especially if its arguments were more intuitive than emtrend's.