Howdy, Stranger!

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

Supported by

GLM with BayesFactor

edited April 2016 in JASP & BayesFactor

Hello, I am looking into using Bayes Factor package for a new experiment. I have a binary dependent variable, and usually, I would analyse my data using a glm or glmm (family="binomial").

I'm having trouble finding an appropriate function or library to computer the BF statistics for model comparisons. Am I overlooking something, or is this really missing/not implemented yet?

Thank you for your help,
Alasdair

Comments

  • EJEJ
    edited 2:28PM

    Hi Alasdair,

    This is not implemented yet! It is on the list though.

    Cheers,
    E.J.

  • edited 2:28PM

    Thanks for your swift reply.

    Do you know if it is implemented in any other packages? I see that Gelman has a bayesian GLM function in his package, but I think it is doing a different thing from the bayesian LMs in your package (ie, it's not computing BF statistics).

    I'm guessing it wouldn't be too hard for me to compute the BF statistic for the GLM by hand (well, using an R script).

  • EJEJ
    edited 2:28PM

    I believe that there's a R package by Raftery that could be helpful (forgot the name though).
    Cheers,
    E.J.

Sign In or Register to comment.