Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

Table of prior distribution

Hi,
I would first like to mention I am very new to JASP and bayesian statistics!

I want to use bayesian statistics in my study to complement the usual null hypothesis testing. I have found several null findings, and when running the data through a repeated measures (mixed) ANOVA in JASP, I get a a BF10 of 0.005, which I interpret as strong evidence for the null hypothesis of no effect (in my case).
I however "just" used default priors (of 0.5) as I wasn't able to calculate an informed prior myself (unsure which articles to base on). What I did instead, was run the analysis a few times with different priors set to: 0.1, 0.25, 0.5, 0.75 and 1.
I made a small table out of this to demonstrate that even with changing priors, the BF10 stays <0.3.
I interpreted this that even if we assume a small or large effect, the null hypothesis of no effect seems most likely true.

Is this a correct/accepted approach? Am I missing something?
Thank you!

Comments

  • Yes that makes perfect sense! (note that the BF says something about the evidence from the data -- if you want the probability of a hypothesis being true then you need to take the prior model plausibility into account). So the safest conclusion is that across a wide range of prior distributions, the data provide evidence for the absence of an effect.
    E.J.

Sign In or Register to comment.