omit intercept lines in plot of ordinal regression from "effects" package in r

user_9777958 0
Viewed 56

I want to visualize a mixed effects ordinal regression, fit using the clmm function within the ordinal package in r. I am using the effects package. However, I do not want the intercepts to be displayed on the plot (thresholds between response categories - seen as dashed horizontal lines below). Any guidance on how I can omit them?

Example code and plot:

library(ordinal)
library(MASS)
library(effects)

mod <- clmm(poverty ~ age + (1|country), data = WVS)
plot(Effect("age", mod, latent = TRUE))

enter image description here

intercept regression effects plot r
1 Answers

It reads off the thresholds stored in your eff object.. Might have some other options in plot.eff, but for now you can write a function to set the threshold to NULL and plot:

plot_without_intercept = function(eff){
       eff$thresholds = NULL
       plot(eff)
}

Eff = Effect("age", mod, latent = TRUE)
plot_without_intercept(Eff)

enter image description here

user_122584590
Related