Hi All,
I am using brms_multiple to include 20 imputed data sets in my model. After running the model the Rhat and ESS seem OK (all <1.04 and >1000). But I do get a warning message:
Warning message:
The displayed Rhat and ESS estimates should not be trusted for brm_multiple models. Please see ?brm_multiple for how to assess convergence of such models.
Why should I not trust my results, I cannot find the anwser online. I am new to the brms package so any advise would really help my project. Thanks in advance!
Hi Celine,
I recommend reading brms’ vignette for missing data imputation. Specifically:
In the summary output, we notice that some Rhat values are higher than 1.1 indicating possible convergence problems. For models based on multiple imputed data sets, this is often a false positive: Chains of different submodels may not overlay each other exactly, since there were fitted to different data… Such non-overlaying chains imply high Rhat values without there actually being any convergence issue. Accordingly, we have to investigate the convergence of the submodels separately.
Thankyou AlesVomacka,
That overview helps a lot!
1 Like