Please help the question. Thank you!
A survey organization wants to take an SRS in order to estimate the proportion of pople who have seen a certain television program. To keep the costs down, they will take as small a sample as possible. But their client will only tolerate chance errors (deviations from the expected value) of 1 percentage point or thereabouts. Should they use a sample of size 100, or 2,500 or 10,000? You may assume that the population is very large and also that the population proportion will be in the range of 20% to 40%.