I'm doing the analysis with a linear mixed model.Because the sample size is different, the Scheffe test is used for the multiple comparison of my data. However, I do not know how to write code in R, and it is suggested that I cannot directly use the data in lme for multiple comparison. How can I solve this problem?
emmeans is indeed easy to use, here's an example of different approaches to get contrasts and pairwise comparisons:
library(lme4)
library(nlme)
library(emmeans)
data("Orthodont")
m
For a mixed model, your best bet is probably to use the emmeans package for multiple comparisons.
emmeans is indeed easy to use, here's an example of different approaches to get contrasts and pairwise comparisons:
library(lme4)
library(nlme)
library(emmeans)
data("Orthodont")
m
Daniel Lüdecke hi, thanks a lot for sharing the example codes, I am a bit confused about em
> as I print the data Orthodont, and all values of age are 8,10,12, 14, not values of 50 and 60.
oh, that might be. I just chose random values, of which I thought that they were in the data. The at-values "50" and "60" (or maybe then "8" and "12") are just two values for age, which you can use for pairwise comparisons of female (age 8), female (age 12), male (age 8) and male (age 12). It might be easier to think of two categorical predictors, where you compare each combination of groups with each other in a pairwise comparison, I can't quite remember why I chose this toy example in my answer.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size