Skip to main content

Table 1 Summary of how to obtain the CI’s of the fixed effects and the variance-covariance parameters

From: A comparative study of R functions for clustered data analysis

 

95% confidence intervals of parameters

 

β

{σ2,ρ}

{σu,σε}

geese

Estimate ± 1.96 standard deviation

gls

confint

intervals(,which=“var-cov”)*

 

intervals(,which=“coef”)

  

lme

intervals(,which=“fixed”)

intervals(,which=“var-cov”)

lmer

confint(,method=“Wald”)

confint(,method=“profile”)

 

confint(,method=“profile”)

  
  1. *Applying the “intervals” function to a “gls” fitted object returns the CI of σ instead of σ2