Due to Underdispersion of my count data, I am considering a generalized Poisson regression to model the effect of 3 independent variables on the dependent count variable. One of these independent variables has a significant effect, and I would like to perform a mean comparison. I am however not successful with the emmeans package in R as this does not support the vglm function. I will appreciate if someone has an idea on how to address this.
Thank you in advance
Please try this in R:
library(lsmeans)
library(multcomp)
> means summary(means)
CLD(means, alpha=0.05, Letters=letters, adjust="tukey")
# generate fake data that might look like what you want ------------------------ set.seed(1839) # for replicability n
Thank you Abdel.
I have tried the script but the error has persisted.
Please here's what I got:
"> means
Thank Mohamed,
With the regular glm, "emmeans" works perfectly as you rightly suggest.
My concern arises when I apply "vglm" instead of "glm" in a situation where the data are underdispersed. I came across "vglm" online but it didn't work for me.
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training