Timeline for influence of bayesian priors: rjags and categorical variables
Current License: CC BY-SA 4.0
8 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 23, 2022 at 19:29 | vote | accept | denis | ||
| Apr 23, 2022 at 17:17 | answer | added | Christian Hennig | timeline score: 4 | |
| Apr 23, 2022 at 16:19 | history | edited | denis | CC BY-SA 4.0 |
added 2425 characters in body
|
| Apr 23, 2022 at 14:07 | history | edited | denis | CC BY-SA 4.0 |
deleted 303 characters in body
|
| Apr 23, 2022 at 14:06 | comment | added | denis | MMmmm ok I think I see. So I will find the frequentist result only for large enough priors (and enough iterations), but the results for narrow priors will be different from the frequentist solution, whatever the number of iterations ? | |
| Apr 23, 2022 at 14:03 | review | Close votes | |||
| Apr 23, 2022 at 14:18 | |||||
| Apr 23, 2022 at 14:03 | comment | added | EdM | Why do you think that the posterior probability sampled by MCMC would be independent of the choice of prior? Your "expected solution" is only the frequentist estimate. It's quite possible that your first Bayesian results with the narrow prior are completely correct for that choice of prior. | |
| Apr 23, 2022 at 13:38 | history | asked | denis | CC BY-SA 4.0 |