Specifying interactions in bayesian linear regression
This seems like it should be obvious, but I don't know how to specify an interaction between two quantitative predictors in bayesian linear regression (or non-bayesian for that matter). I suppose I could just create a product variable, but that seems a bit odd to have to do that for every interaction I want to estimate. Is this feature available?
Comments
Hi dustinfife
In the model specification tool, you can select the variables you want to have interact, and drag both over to the box simultaneously. This will create the desired interaction term. Let me know whether it works.
Cheers,
E.J.
No luck. I clicked two and dragged and got this:
It only shows main effects models. And here's the whole window:
Is this helpful: https://jasp-stats.org/wp-content/uploads/2018/06/interaction_effect.gif
VERY helpful. Thanks EJ!
(Great program, by the way. I can now teach my graduates to do Bayesian without them flipping out
)