My research about Age-Period-Cohort models that is assumed binomial distribution. In this model, there is relationship which connects the three factors, cohort=periode-age(c=p-a).There are different solution to that problem, and the one proposed by Holford(1983) consists of firs adjusting the model with any parametrization of the factors, and later carrying out a regression of each set of estimations over its corresponding factor.

Similar questions and discussions