Mengmeng Ji posted on Wednesday, March 25, 2020 - 4:36 pm
I am wondering if it's possible to run GxE model for ordinal dependent (Phenotype) with a continuous moderator. I found an example code for continuous data. When I tried to modify the P to an ordinal variable, there is a fatal error saying "THIS PARAMETER RESTRICTION MODEL IS NOT AVAILABLE.".
So for continuous variable, we label the variance of twin1 and twin2, then decompose them in model constraint depending on M. But how do we calculate the variance for an ordinal variable?