[R-sig-ME] comparison bet. mixed and ols model (using lme4)

Simon Harmel @|m@h@rme| @end|ng |rom gm@||@com
Wed Sep 16 18:41:19 CEST 2020

Hello all,

In my regular, non-mixed-effects model below (`ols2`), I have just used the
average of each cluster as DV and used a binary cluster-level predictor
(sector) in the model.

# Questions: Should the amount of total variation in `ols2` model equal the
between-cluster variation (tau_00) in a corresponding mixed-effects model

Thank you, Simon


hsb <- read.csv('
hsb_ave <- hsb %>% group_by(sch.id) %>% mutate(math_ave = mean(math)) %>%

ols2 <- lm(math_ave ~ sector, data = hsb_ave)

m1 <- lmer(math ~ sector + (1|sch.id), data = hsb)

	[[alternative HTML version deleted]]

More information about the R-sig-mixed-models mailing list