Question: A recent study indicated that 29% of 100 women over age 55 in the study were widows.
a) How large a sample must one take to be 90% confident that the estimate is within 0.05 of the true proportion?
b) If no estimate of the sample proportion is available, how large a sample is needed?