The problem I want to solve:

Lets imagine that I have two factories A and B, where each factory produces coins. What I suspect is that the probability of tails (denoted as $\theta$) varies substantially between the two factories.

How is my data:

My observations are 10 draws for each of the three coins in factory A and another 10 draws for each of three coins in factory B

How I would model the problem: I would use a beta binomial hierarchical model with the following form:

For coin $j$ in factory A the sequence of 10 draws comes from a binomial distribution of the form $yi∼Binomial(N=10,\theta_j)$, and then $\theta_j$ values depend on a beta distribution $\theta_j∼Beta(α,β)$ (which I suspect that it is different for A and B factories).

What is my problem

On the different textbooks I have check ( Gelman et al BDA3 and Kruschke DBDA 2014) the hierarchical models are used for doing inferences on the $\theta_j$ parameter of the binomial distributon. However, in this case what I wanna check is if the parameters $\alpha$ and $\beta$ of the beta distribution is different between the two factories.

My question

Finally, as indicated above, what I want to know is if the $Beta$ distribution for factory $A$ and factory $B$ differs.

Thanks for reading!!


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.