Learn Bayes suggestions about jasplearnbayes HOT 6 OPEN

Thanks for the suggestions @JohnnyDoorn
(1) About the "data" coming first -- this is consistent with the rest of JASP, right? We usually start an analysis by dragging the relevant variables to the input boxes, and then select the analyses.
(2) The required plot would go under "testing" I assume. Then it would be relevant only for a point H0 versus a composite H1. So it would be an option that could be activated only when these model specification conditions are met. I do think that could make sense. However, in terms of priorities I hope we can have a z-test first, followed by a directional test.

from jasplearnbayes.

I wouldn't mind having the specifications of models and hypothesis before the data. That agrees with the model and hypothesis first principle. It will indeed a bit different from standard analyses such as the z-test, but there default choices are made. When learning Bayes I think that first putting forward hypotheses is quite reasonable.

from jasplearnbayes.

There is not a single analysis in JASP where we put data second, right? The data always come first in JASP. Also in Summary Stats.

from jasplearnbayes.

In the audit module data comes later. In the standard analyses the data indeed come first, but that's because the models and default priors are chosen once you've chosen the associated analysis. In the learn Bayes module you don't only study default analyses. Considering the flow of information from prior, using data to update to posteriors. From a teaching perspective I think it's reasonable to specify priors first.

from jasplearnbayes.

OK, I see. Well I can be pursuaded. @FBartos what do you think?

from jasplearnbayes.

hmm, I don't have a strong emotional attachment to any order in particular, so I wouldn't mind changing it.

from jasplearnbayes.