Brms Coef graphing

Interfaces brms
1

Hey all,

I am a research assistant currently working on a project using brms. I had a question about achieving something specific in graphing my results, namely, I have an ordinal regression model that is using the cumulative family. I have a grouping variable for intercepts/breakpoints and I was hoping to generate a plot similar to this quick plot generated with the depreciated package brmstools as seen here:

Rplot2

However, I would like to be able to see the differing intercepts based on the grouping variables neatly stacked upon one another. Does have any insight into how I might go about doing that or perhaps where I should look? I am having trouble finding a clear cut answer. Apologies if this is a rather simple thing that’s been answered elsewhere. Thanks for the help in advance!

  • Jeremy
  • Operating System: macOS Mojave 10.14.5
  • brms Version: 2.9.0
2

If you are using the cumulative category and have something like (1 | group), then you only get one parameter per participant–the amount that all the thresholds are shifted for a given group. You can easily recover those parameters using coef(), but the tidybayes package will make it very easy to get the values and post-process/plot the results.