I would like to check for publication bias on a BRM object. has any idea?
What do you mean by checking a BRM object for publication bias?
What I’ve seen is that you can use meta-analysis and inspect the funnel plot and other statistics to see if a phenomenon of interest has evidence of publication bias. I think that was what you’ve meant? There is a package called Bayesian Meta-Analysis via Stan • MetaStan which can help you. Also, @Solomon has a nice collection of posts explaining how to perform meta-analysis: Bayesian meta-analysis in brms | A. Solomon Kurz and Bayesian meta-analysis in brms-II | A. Solomon Kurz.
I run a meta-regression model with brms library (Bayesian regression models using STAN),
Can i use the Egger’s test? because I saw that it is used for non Bayesian meta-analysis
I run a meta-regression model with brms library (Bayesian regression models using STAN),
Can i use the Egger’s test? because I saw that it is used for non Bayesian meta-analysis
Hi,
unfortunately, frequentist tests don’t translate directly into Bayesian framework. Is there something specific in the tutorials @storopoli linked to that you find problematic? If you are having trouble understanding what the results mean, you migh want to consult some of the resources linked at Understanding basics of Bayesian statistics and modelling. Additionally, Hypothesis testing, model selection, model comparison - some thoughts has my current thoughts on some possible approaches to map scientific questions to Bayesian models.