Bayesian mediation
Sorry for the offtop(not about JASP directly) but I don't have many specialists to consult with, and maybe you, fine people, could share your opinion
The thing:
Bayesian mediation analysis with a sample of 54 observations(can't be enlarged), 1 dependent (binary) , 1 binary factor/mediator and 4 non-binary variables (standardized). No prior knowledge, so uninformative priors are: "default" normal and recommended by Gelman, et.al. 2008, "A weakly informative default prior distribution for logistic and other regression models." - Cauchy (to compare) Both at 0, 2.5. But instead of intercept of 0,10 i pick 0, 5 just because it seemed less wide for such data. And I didn’t find proper info to justify this—is it ok or this is a flaw?
Comments
Seems reasonable, but when in doubt conduct a sensitivity analysis
EJ