[R-sig-ME] Marginal effects plots with interactions of glmm model averaging

Vitor Vieira Vasconcelos v|tor@v@v @end|ng |rom gm@||@com
Sun Mar 21 17:11:49 CET 2021


Good morning!

    I was guessing if it would be possible to make marginal effects plots
with interaction terms (two way, three way, etc.) of model averaging
results of glmm, similar to interaction plots of a single glmm, such as in
the ggeffects, effects and sjPlot packages. I saw in the interplot package
that it would be possible to make these "ad hoc" effect plots if we provide
a dataframe with four columns: the scale of the conditioning variable, the
simulated interactive effect at each break of the conditioning variable,
and the simulated lower bound and upper bound of the confidence interval.
However, I am not sure about how could I calculate these values for a model
averaging. Right now I am using Mumin package for the model averaging of
glmm.
   I would be very grateful if anyone here could provide comments if this
approach would be possible or advisable, or if you happen to have an idea
about how to operationalize it.

Best regards,Vitor

	[[alternative HTML version deleted]]



More information about the R-sig-mixed-models mailing list