Howdy, Stranger!

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

Supported by

Bayesian Linear Regression -- No categorical variables

Hi,

I am trying to run a Bayesian regression in JASP 0.9.0.0. but the program does not allow adding categorical variables to my covariates -- only continuous ones. Anyone knows why?

Thanks!

Comments

  • It is because the coefficients for categorical variables require a different prior structure, and this is still being sorted out in the BAS package. What you could do is use the Bayesian "ANOVA" (really a linear mixed effect model) and define an ANCOVA.
    Cheers,
    E.J.

  • Hi,

    I am trying as well to run a Bayesian regression in Jasp (0.9.2) using categorical covariates. I tried to follow the suggestion of doing a Bayesian ANCOVA but I am not sure I've done it right. So basically, I have selected "bayesian ANCOVA" in Jasp and included my continuous dependent variable in the relevant box, and my continuous predictor and covariates (one categorical and one continuous) have been included in the covariate box, without adding fixed or random factors. Could this work? Also, is it possible to have the posterior and prior probability plot with mixed ANOVA/ANCOVA?.

    Thanks!

  • Yes, that could work. The next version of JASP will have posterior distributions for the ANOVA/ANCOVA.

    Cheers,

    E.J.

Sign In or Register to comment.