se of variance/correlation estimates

[Ron Crump <rcrump@metz.une.edu.au>]

Hi,

I have an animal model analysis in which
one of the genetic correlation estimates
gets fixed at 0.999 (CORR used, with !GP)
during the course of the analysis. The
other genetic correlations are reported with
negative standard errors at the end. Does
this mean my analysis has failed, if so will
fixing the high genetic correlation from the
start fix it? Why have I got negative se estimates?

Ron.