A study is conducted to estimate the average difference in bus ridership for a large city during the morning and afternoon rush hours. The transit authority's researcher randomly selects nine buses because of the variety of routes they represent. On a given day the number of riders on each bus is counted at 7:45 A.M. and at 4:45 P.M., with the following results.

Bus Morning Afternoon

1 44 41

2 51 47

3 37 44

4 24 32

5 50 46

6 44 39

7 50 47

8 55 51

9 48 49

Use the data to compute a 90% confidence interval to estimate the population average difference. Assume ridership is normally distributed.