In JMP help under Discrete Fit Distributions, the following is mentioned:
Beta Binomial
This distribution is useful when the data is a combination of several Binomial(p) distributions, each with a different p. One example is the overall number of defects combined from multiple manufacturing lines, when the mean number of defects (p) varies between the lines.
Does anyone has a reference, article, supporting this statement?
Kind regards
On this page:
https://www.jmp.com/support/help/en/15.2/index.shtml#page/jmp/discrete-fit-distributions.shtml#ww1162265
Agresti and Coull are referenced here.
Agresti, A., and Coull, B. A. (1998). “Approximate is Better Than ‘Exact’ for Interval Estimation of Binomial Proportions.” American Statistician 52:119–126.
