The purpose of this problem is to interpret a confidence level correctly while avoiding misconceptions. To estimate the average salary among people in a certain district, a marketing team obtains a simple random sample of 100 people from that district. The team sets up the following approximate 90% confidence interval, using a normal estimate, for the unknown average salary (in dollars annually) of the district: [41500, 49750] Approximately 90% of the people in the district have a salary in the interval (41500,49750). The procedure used to construct the interval works approximately 90% of the time. For the normal estimate of 90% confidence to apply, the salary distribution in the population must be approximately normal.