Hello everybody, I am fitting a multiple outcome mixed effects model where the correlation between the outcomes are going to capture by the random effects (random intercept). The random intercept follows N~(0,sigma_u) I am having a hard time of selecting an appropriate prior for the sigma_u. I …

As Erling said, your priors are likely causing a lot of problems here. With the function inv_logit(x), as soon as x gets larger than ~19, the function will only return 1. So you should look at making those priors significantly smaller. Additionally, you should try the non-centered parameterisation …

Looking for some advice based the n_eff and the pairs plot for mixed effects models

General

Ara_Winter September 17, 2020, 12:20am 2

Welcome!

Over here Reparameterizing to avoid low E-BFMI warning this might give you some pointers.

2 Likes

Topic		Replies	Views
How can I solve "bfmi-low" problem? Modeling	22	11333	April 20, 2018
Reparameterizing to avoid low E-BFMI warning Modeling fitting-issues	7	4665	October 2, 2022
Multilevel model implementation - tricks to solve low neff, low E-BFMI, mildly high Rhat? Modeling techniques , fitting-issues , specification	15	1233	June 26, 2023
Solving BFMI issue - could anyone suggest a better prior? Modeling fitting-issues	1	465	September 6, 2020
Simple intercept only model with poor chain mixing Modeling rstan	6	591	September 17, 2021

Looking for some advice based the n_eff and the pairs plot for mixed effects models

Related topics