This addresses #14 by replacing the "Other distributions" section with an example based on the same real-world data as the preceding sections. The idea (in terms of instruction) is to show the need for a different probability distribution by checking the model residuals, thereby addressing also #13. A quick check on basis dimension is also included (addresses #12). The old section on "other distributions" has been moved to the end of the presentation.
Fitting method for all models after section 2 has been changed to REML, in accordance with #10.
This branch does not change much in the section on mixed models, since this part requires a more substantial overhaul (#17).
This addresses #14 by replacing the "Other distributions" section with an example based on the same real-world data as the preceding sections. The idea (in terms of instruction) is to show the need for a different probability distribution by checking the model residuals, thereby addressing also #13. A quick check on basis dimension is also included (addresses #12). The old section on "other distributions" has been moved to the end of the presentation.
Fitting method for all models after section 2 has been changed to REML, in accordance with #10.
This branch does not change much in the section on mixed models, since this part requires a more substantial overhaul (#17).