What does check_model() do for mixed beta regression?

[tbx200]

I'm fitting a mixed beta-regression model in glmmTMB.
After reading some literature (Espinheira et al. 2008 & Ferrari & Cribari-Neto 2004), I'm wondering what exactly is calculated and plotted for this type of model. I understand that model diagnostics for this type of model have not been fully crystalized yet.

Are the calculated residuals, weighted residuals (and which type), or are they standardized ordinary residuals?

[bwiernik]

It uses the standard residuals() function from the glmmTMB package. The default for this function is "response", that is, simply res = response - fitted residuals, not weights or scaling. There is potentially an argument that we should use weighted Pearson residuals instead, but currently the default response residuals are used.