Hello all!
I appreciate any guidance on using Bayesian modeling where there are between-site measurement differences. Is there any approach using rather brms and if not, in Stan?
I considered using one of the measurements unrelated to other measurements yet very site-specific as an instrumental variable.
I would appreciate it if sharing your insights.
Sorry nobody got to this sooner. Hope it’ snot too late.
You can build a hierarchical measurement error model, assuming that you mean the error distributions are different. There’s a high-level overview of measurement error models in the Stan User’s Guide:
https://mc-stan.org/docs/stan-users-guide/measurement-error.html#bayesian-measurement-error-model
@andrewgelman and I have a very simple example for Covid PCR test sensitivity and specificity here:
1 Like
@Bob_Carpenter Thank you very much for your kind guidance. Very helpful examples. I am still working on this so it is not late and a very on-time response. I appreciate it.