Multiple comparison correction in plot_model_comparison

[JasperVanDenBosch]

Currently all the MCC happens in the visualization code and not on the Result object. Maybe we could do this on the Result object / consolidate the code. The default options should probably be similar to avoid confusion as to differences between the plot and the Result object

[jdiedrichsen]

As argued - my strong preference for transparancy is to internally store and use uncorrected p-values.
For plots and high-level inference routines, MCC should be an option that can be switched on.

Automatic MCC is confusing and doesn't make sense from a scientific standpoint.

Joern

[JasperVanDenBosch]

Agreed - the default could be False for both - as long as they share the same default behavior or at least make the behavior obvious to prevent this kind of confusion