Suppose that, in Example 3.3, the conditional distribution of i given a is binomial ( i , i )…

Suppose that, in Example 3.3, the conditional distribution of i given _ is binomial (i, i) instead of Bernoulli, where i is a known positive integer, and i is the same as in Example 3.3. Show that, with a suitable link function, this is a special case of GLMM. What is the dispersion parameter _ in this case, and what is ai(_)?