How to explain beta binomial prior for regression analyses?
Im doing some regressions, with a total of 25 models involved. Assuming a uniform prior is simple in JASP. However, I want to specify other priors, for example so that one of my models get a .25 prior (and the other 24 models thus sum to .75). This can be done when playing around with the Beta binomial distribution that is by default set as a=1 and b=1. But rather than writing in my article that "after playing around in JASP with the settings for the beta binomial distribution of the prior I managed to set the priors so that model X had a prior of .25 and the other models..." I want to state why it makes sense to derive my prior distributions using the beta binomial distribution
So can someone explain in somewhat plain english what one does when assuming the beta binomial distribution under various settings?