search

jasp-stats / jasp-issues

This repository is solely meant for reporting of bugs, feature requests and other issues in JASP.
58 stars 29 forks source link

Request: Robustness check/different priors for BANOVAs #201

Open kevinvschie opened 6 years ago

kevinvschie commented 6 years ago

Template for enhancement requests

I would like to request the possibility to check the robustness of Bayesian ANOVAs. I know this feature is already available for t-tests, but it would be great to have it for Bayesian ANOVAs too. I have had this as a reviewer comment so many times. A similar request would be for the sequential analysis. The request applies for ANOVAs with within as well as between subjects factors.

tomtomme commented 7 months ago

related: https://github.com/jasp-stats/jasp-issues/issues/1719

tomtomme commented 7 months ago

Important remarks from the duplicate #2319

[JohnnyDoorn] Creating those plots would result in very long run-times. For now it seems that the only possibility is to do this manually, and set the prior parameters to different values yourself. It could be an option for JASP to at some point offer a checkbox that reruns the analysis for 2 or 3 other value of the prior distribution.

[EJWagenmakers] Also, there are multiple priors in the ANOVA, so we'd have to think about which ones to vary.