Hi, In this question when we are evaluating the degree of freedom we have just taken the count of age groups as the degree of freedom (df). However since the underlying mortality experience is from a standard table should we not be deducting 2 from the df... Kindly guide. With regards, Pratik
If the question was about testing the goodness-of-fit of a set of graduated mortality rates (where the graduation had been performed with reference to a standard table), then the number of degrees of freedom (df), would be the number of age groups, minus the number of parameters estimated in the formula used to calculate the graduated rates, minus some further df (eg 1 or 2) because of the constraints imposed by the choice of the standard table. But this question is not about testing graduated rates. The word "graduation" doesn't appear at all. This question is about comparing two independent data sets, so we just use the number of age groups compared as the number of df in the chi-squared test. I'd recommend you have another look at Section 5 of Chapter 13. This question relates to Section 5.1.