Question 2ii, April 2020 says that offsetting beside the use in removing aliasing, it has other uses. Can anyone explain in more detail how this happen? As I understand, offset term (xi) added to Linear predictor as below: LP = X*beta + xi How the offset term (xi) can help with the - aliasing, - NCD, - (remove) the effect of an regulatory restricted factor? (as stated in the answer) Thanks very much,
Offsetting in discussed in Section 6.5 of Chapter 16 - have another read of that. Basically, it fixes the effect of one variable (eg NCD or gender if regulations prohibit you from rating on that) and then adjusts all of the other relatives of the other factors accordingly. If you want to know more than this, then perhaps do a google search.