I am using the Design library and robcov to compute variance-covariance matrices for clustered data regression. Is there an easy way to compute the F-test (i.e. linear hypothesis) for clustered data regression ? Thanks in advance! Benn