Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

10*SD uniform hypersd prior fails with small effect sizes / small SDs #124

Open
wwiecek opened this issue Dec 28, 2021 · 2 comments
Open
Milestone

Comments

@wwiecek
Copy link
Owner

wwiecek commented Dec 28, 2021

Try this, for example:

library(baggr)
K <- 10
df <- data.frame(tau = rnorm(K, 0, .001), se = rep(1, K))
#It's not about SEs:
#df <- data.frame(tau = rnorm(K, 0, .01), se = rep(.001, K))
baggr(df)

HyperSD can be estimated but only with a wider uniform prior.

Setting SD of tau higher fixes this issue

@wwiecek wwiecek added this to the v0.7 milestone Feb 24, 2022
@wwiecek wwiecek changed the title Default tniform hypersd prior fails with small effect sizes / small SEs Default uniform hypersd prior fails with small effect sizes / small SEs Apr 25, 2022
@wwiecek
Copy link
Owner Author

wwiecek commented Apr 25, 2022

You can also experiment with increasing/decreasing SD on tau and then playing with x in prior_hypersd = uniform(0, x). You'll see that at smaller scales it has a problem

@wwiecek wwiecek changed the title Default uniform hypersd prior fails with small effect sizes / small SEs 10*SD uniform hypersd prior fails with small effect sizes / small SEs Apr 25, 2022
@wwiecek wwiecek changed the title 10*SD uniform hypersd prior fails with small effect sizes / small SEs 10*SD uniform hypersd prior fails with small effect sizes / small SDs Apr 25, 2022
@wwiecek
Copy link
Owner Author

wwiecek commented Apr 25, 2022

Resolving this should probably wait for when we are deciding on new default priors. See #139

@wwiecek wwiecek modified the milestones: v0.7, v0.8 Dec 18, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant