Maybe we can integrate with #574

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 57.50%. Comparing base (4d523a7) to head (24dd049).
Report is 7 commits behind head on main.

@@            Coverage Diff             @@
##             main     #575      +/-   ##
==========================================
+ Coverage   57.30%   57.50%   +0.19%     
==========================================
  Files          87       87              
  Lines        6416     6436      +20     
==========================================
+ Hits         3677     3701      +24     
+ Misses       2739     2735       -4     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

What is not quite clear to me is the demonstration of the relationship between p and pd? Your introduction ends with

Many of these assumptions can be checked using the check_model() function. We will demonstrate this function using the mtcars dataset.

The reader now knows why model diagnostic is important, why eye-balling is preferred over statistical tests, which kind of assumptions are important to test, and how to do this (using check_model()).

But then you start with simulations to show the relation between p and pd, and that interpreting p-values is not easy. Why this section? This is not directly related to model diagnostics (regarding show cases of check_model()).