Non-parametric Bayesian Mixed ANOVA Or Bayesian Linear mixed Model
Hello,
I have some data that still appears non-normal even after Log and Sq root transformation. Is there anything I can do to do some sort of non-parametric mixed ANOVA in JASP or in R? I've seen some say frequentist linear mixed models are better with non-parametric data and I see the new release of JASP has a bayesian linear mixed model function - Would that help? A bit confused as I'd seen previously that the Bayesian Mixed ANOVA was already a linear mixed model? Any tips would be greatly appreciated!
All the best,
Gabriel
Comments
Hi Gabriel,
Yes, the Bayesian ANOVA is a mixed model but we didn't accommodate the long format nor did we add all the bells and whistles at the time. About the non-normality: I guess there are the Box-Cox transformations, have you considered these? If the data are still non-normal after the transformations you tried, this may suggest that there are outliers.
A rank-based procedure is a good idea, but I am not aware of those procedures for a mixed model....
E.J.