dummy predictor variables in bayesian linear regression
hello jaspers,
I did a Bayesian independent t test and tried the same data set using Bayesian linear regression. After I converted the two groups to a dummy variable with values 0 and 1, I tried it on Bayesian linear regression. however, the dummy variable could not be transferred into the covariate box.
I did the same process using classical independent t test and classical linear regression and there was no problem.
I wonder if Bayesian linear regression cannot be used if the predictors are categorical. I would appreciate any help on this. thanks a lot
cheers,
narcilili
Comments
Hi Narcilili,
"I wonder if Bayesian linear regression cannot be used if the predictors are categorical." Yes, this is correct. Currently the prior structure is set up so that the results are invariant under changes of the measurement scale, but that is not required or desirable for categorical predictors. With categorical predictors, we are in ANCOVA territory. But ideally we'd solve this behind the scenes -- a project for the future.
Cheers,
E.J.
Hello E.J.,
Thank you for the quick reply and explanation. I look forward to future projects in JASP.
Cheers,
Narcilili