[R-sig-ME] Interpreting variances in lme with varIdent weights

[S Ellison]

 We've managed to get ourselves confused in interpreting lme's output and would appreciate clarification; I and a colleague differ on interpreting the words used in lme's help pages and output, and I see his point; it can be read two ways.

Using varIdent weights, lme provides a factor for each varIdent stratum. The question is whether that factor applies to standard deviations for _all_ random effects or just to the sd for the residual term within each stratum.

Any guidance appreciated.

S Ellison
LGC

*******************************************************************
This email and any attachments are confidential. Any use...{{dropped:8}}