Message/Author |
|
|
Hello, I wondered how best to model an outcome variable that has a bimodal distrubution in a path model. The predictors in the data set are categorical and continuous. Of the continuous scores two are positively skewed and have been corrected using transformations. Thanks, Ellen |
|
|
I'm not sure. You could create an ordered categorical variable. I would ask this on a general discussion forum like SEMNET. |
|
|
Thanks so much for your response. I will submit my question there also. I wondered about using MLR because I had read on earlier boards that it was better for non-normal continuous variables. Thanks, Ellen |
|
|
MLR is a good choice for skewed variables but not for bimodal. |
|
Back to top |