Displaying 1 result from an estimated 1 matches for "sigma_beta".
2003 Mar 30
1
simple test of lme, questions on DF corrections
...or the O(p/n) bias of
standard maximum-likelihood estimates. Does anyone know of any
documentation or references regarding this?
Finally, if the within-group rms error sigma and the
between-group rms error sigma_b have been estimated, then a
standard calculation seems to show that the uncertainty
sigma_beta in the estimated population mean beta is:
sigma_beta^2 = sigma^2/N + sigma_b^2/N_groups
The value of sigma_beta reported by lme is 3% higher than this
formula. I could imagine that a DF correction to a ML estimate
gives small differences from this formula for some reason, but
I''d like t...