Hi, Can anyone explain on the last para - why and how is the log of exposure as the offset term and the model form as identical for the model where number of claims is the Y? Thanks.
Page 22 of Duncan Anderson's paper, and the early chapters of McCullagh and Nelder, go into this in more detail, as the Core Reading only requires a limited awareness of the 'weights' used within the GLM model. I would hope that just an awareness of the equivalence is sufficient for the exam.