Howdy, Stranger!

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

Supported by

When to set priors or use default

I have recently run an experiment and shown an effect for group A on X. I would like to run the same experiment in group B to see if we get a different effect on X. Should I use default priors or set them according to the results of the first experiment?


  • I think, if You compare group leave default or set the same priors in both.

  • You could analyse both A and B together, and see whether you find an interaction in an ANOVA. The approach you propose (i.e., the use the posterior from A as a prior for the analysis of B) is appropriate if you want to test the hypothesis that B shows no effect whatsoever. But you indicated you want to know whether the effect in A is just as large as in B, and this calls for a test of an interaction.

Sign In or Register to comment.