Bayesian A-optimal design for linear models

[ckogan]

I am interested in constructing an A-optimal design for a Bayesian linear model with a diagonal error covariance matrix where the error variance depends on one of the categorical variables in the analysis. I was wondering if and how SAS proc optex could be used to construct an optimal design for this problem?

Thanks,

[Rick_SAS]

Does this example offer any help?

Example 14.6 Bayesian Optimal Design SAS/QC(R) 13.2 User's Guide

along with the statement  GENERATE CRITERION=A;