Two populations are surveyed using simple random samples. the populations are large enough compared to the samples that we can ignore the finite population correction factors. a sample of size n is used for the first population, which has a population sd of s1. a sample of size m = 2n is used for the second population, which has a population sd of s2=2s1. in which of these two samples would you expect the estimate of the population mean to be more accurate?

the answer is trees if i'm not mistaken

"1. it should be, it has been for about 20 years, and it's really not much different in the end than breeding animals to have particular traits. there are always concerns about genetically engineered animals/plants affecting the environment in unexpected negative ways, but i can't think of an example where the fears didn't turn out to be baseless.  

first, as you point out, it would be a great advance in pharmacology, to be able to tailor drugs and their doses to me. the pros involve better medical treatment. the cons involve privacy issues like insurance companies denying coverage to people with genetic risk factors. looking at dna markers wouldn't affect the frequency of genetic diseases, but it would allow detection of them earlier. and ok, actually i suppose that if people used this information to avoid having children with genetic diseases, some might decrease in frequency."