Howdy, Stranger!

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

Supported by

BF = Inf How is this possible?

I've just conducted a Bayesian repeated mesures ANOVA.

In the model averaging table (The output you get when you click the 'effects' button), it says that the BF for one of my main effects is infinite.

How can this be possible, considering that even if the alternative vastly out-weighs the null, there should still be a very very (maybe extermely) small posterior for the null? In which case, the BF should always be some finite (albeit maybe very large) value (except maybe when the prior contains 0 values?).

Is this just a case of JASP saying that it can't handle extremely large values, or is there something I am missing in my understanding?

Comments

  • Can you post a screen grab of the models and effects tables?

  • Yes, it's an issue of numerical precision. Apparently the posterior mass of the models *excluding* that main effect sums to epsilon, with epsilon so close to zero that the number cannot be represented on your computer.

    E.J.

Sign In or Register to comment.