Is there a way that I can run a joint multivariate model that allows all intercepts and slopes to covary but does not allow level-1 residuals to covary across variables.   I am using SAS OnDemand for Academics and the following code will not run.   proc mixed data=homework3 method=ML noclprint; class school dv; model popst= conss const sexs sext/ noint s; random conss sexs sext const/ sub=school type=un; repeated dv / subject=school*pupil type= UN(1); run; ... View more