Dear OpenMx team and users,
I am running an AE multivariate twin model with a sample of 21 DZ twin pairs and 50 MZ twin pairs on 4 variables. The Genetic correlation between two of these variables is -1.10.
So my obvious question is; what possible reasons could explain this correlation bellow -1? Could it be related to the absence of "C" or "D" in the model? Or could it be related to the sample size?
Many thanks,
Arthur
Printer-friendly version
For some choices of model parameterization, a genetic correlation outside [-1,1] is certainly possible.
Yes, your sample size is quite modest, especially for an analysis of 4 phenotypes, so all of your parameter estimates will be subject to a lot of sampling error.
Thank you very much for your reply!