Suppose we want to study about the percentage of consumers who are going to purchase a brand extension next year

By secondary sources we already know the distribution of the consumption variable, which is going to be used to split the population:

Heavy users: 25%

Light users: 75%

a) If we consider a population of 70,000 consumers and we apply a stratified sampling, how large a sample do they need in the first group (heavy users) at a 95.46% level of confidence (z=2) and a sampling error of +/- 4%?

b) Once the study is finished, the percentage of consumers who would purchase this brand extension for each level was:

Heavy users: 65%

Light users: 35%

Which is the total percentage of consumers for this new brand extension?

c)With this information, which is the real sampling errorat a 95% level of confidence?

If you submit the solution for this third question you may receive an extra point