Consider iid random variables X1,...Xn of mothers' ages in ears arising from a population with mean u1 and variance o1. Corresponding to Xi, is the age Yi, in years of the mother's eldest child. The ages of eldest children who have mothers that are still alive arise from a distribution with mean u2 and variance o2. There is interest in testing H0: u1 - u2 = 20 versus H1: u1 - u2 > 20. Approximately how large should n be such that the power of the test is 0.90 when u1 - u2 = 25 years? you can assume that the standard deviation of the age difference is 10. You can assume that the significance level is 0.05

I really need help in this question. Any help would be appreciated.