I have a stratafied sample, with only two strata. The sample is stratafied because it is thought that the stratafication variable may be correlated with the outcome. It's also probably safe to assume that there is some degree of homogeneity within strata.
Because of this homogeneity within strata, is it acceptable to use the CLUSTER option in the VARIABLE command? That is my code would read, CLUSTER IS STRATAFICATION VARIABLE.
I know that the strata isn't technically a cluster, but it seems to me that they can share the important characteristic of correlated errors. Am I nuts?