Thank you for the suggestions. All of the models I am running contain the same suite of variables, however the variable show differing predictive performance between the models. I am not currently looking to perform variable selection across each model as I do not have access to the computing power required.
My biggest issue is currently in determining whether a predictor is having an appreciable effect on my response. I can eyeball this by examining my pots, however I am looking for a more concrete way of determining importance.
As an example the below plot shows very little meaningful information:
However a much more meaningful trend can be observed in this plot:
Is there any way of drawing a line between these two in brms? If I was using mgcv I would rely on the p-values reported for each term, however I am unsure of an equivalent way in brms.
Thanks again.