You are interested in seeing if a new Facebook promotion increases sales by 25% or more. The promotion that you are using now gives sales of about 10%. You would like to run a test to see if a new promotion improves sales. You would like the test to find a significant result if the new sales percentage meets the mark given here 85% of the time. Assume a type 1 error rate of a=0.05.

1. Give the assumptions for this situation

2. Find the sample size in this situation

3. Find the sample size needed if we would like a 50% improvement in the sales numbers

4. Now your boss states that the error rates that we are using are too low. How would you explain that increasing these would be a bad idea.