My data had to be transformed in order to eliminate outliers. So the dependent variable now has several cases with negative values. I cannot run a Generalized Linear Mixed Model because of this. If I move the scale, the outlier issue returns. As you probably guessed, I am using scale, ordinal, and nominal types of data, so I am basically stuck with GLMM. Well, that's what I think, unless there is another option for such a mixed data. Ideas to solve the issue or to use another model would be greatly appreciated:

Oscar