Hey, I've been solving chapter 11. Could i gain a better understanding of when normal distribution is assumed for the population and when it is not? Thanks
first see that distribution of each X,say, is independent of each other. then see the mean is big enough(because then most of the data be symmetric), thumb rule >5. find constant variance. then apply Normal Approximation.......... (& for calculating probability apply continuity correction wherever necessary.
me! I posted as general(Binomial,Poisson, Gamma.....) mean or sum, just about when to use Normal assumption..... am I missing some points to consider?
CI for mean(s) require normal assumption for small samples. CI for variance(s) always require normal assumption. Approximate CI for Poisson/binomial require large samples.