Testing replication with JASP
Hi,
I am interested in determining the success or, more likely, the failure of a replication study using bayesian statistics.
In a previous post you wrote: "sometimes you might want to center the prior on a nonzero value. For instance, this is what we do for the "replication Bayes factor" (http://www.ejwagenmakers.com/2014/VerhagenWagenmakers2014.pdf). We are currently working to add these kinds of tests to JASP".
When do you expect this to be available?
If these features are not yet available in JASP 7.5, what would be the best way to test a replication with the current possibilities?
Best, Marieke
Comments
Hi Marieke,
I think it might take three months or so before this functionality is in. I will take it up with those responsible today. Note: the Verhagen paper does have R code, and so should the other replication BF papers we did (one on correlations: http://www.ejwagenmakers.com/submitted/BayesianBathingHabits.pdf; and one on the "pipeline project": http://www.ejwagenmakers.com/inpress/PipelineProjectInPress.pdf).
Cheers,
E.J.
Hi
When looking for information on using Baysian statistics to determine succes of replication (as suggested in Verhagen & Wagenmakers, 2014) I came across the previous posts. I was wondering if the feature (center the prior to non-zero value) is now available in JASP and if there are some instructions on how to use this feature. Also, I was wondering wether this feature is available for other tests than t-tests or correlations, specifically for ANOVA or ANCOVA.
Thanks!
Anne
Hi Anne,
Quentin Gronau has just done the math for the t-test, and we're looking to submit the paper soonish. For ANOVA this is more difficult, but I did just come across this paper that may be relevant....let me look it up: https://arxiv.org/abs/1611.09341
Not yet available in JASP though
Cheers,
E.J.