Re: [R] nlme::lme sigma parameter standard deviation or variance?

Modifying the first example in help('lme', pac='nlme'): library(nlme) fm1 <- lme(distance ~ age, data = Orthodont) # random is ~ age fm1$sigma fm1.2 <- lme(2*distance ~ age, data = Orthodont) # random is ~ age fm1.2$sigma Conclusion: Standard deviation, as

Re: [R] nlme::lme sigma parameter standard deviation or variance?

You can run a test. Multiply all your data by a scalar, say 2. If this changes the result lme_mod$sigma by a factor of 2, then it is a std deviation. If it changes the result by a factor of 4, then it is a variance. HTH, Eric On Sat, Dec 18, 2021 at 11:26 AM Courtney Van Den elzen wrote: > > Hi

[R] nlme::lme sigma parameter standard deviation or variance?

Hi R-help, I am a researcher fitting linear mixed models using the package nlme. I am wondering whether the sigma value that is outputted as part of the model object is standard deviation or variance? for example, I might fit a model lme_mod <- nlme::lme(response ~ predictor1 + predictor2,