A group of high-risk automobile drivers (with three moving violations in one year) are required, according to random assignment, either to attend a traffic school or to perform supervised volunteer work. During the subsequent five-year period, these same drivers were cited for the following number of moving violations:

NUMBER OF MOVING VIOLATIONS

TRAFFIC SCHOOL	VOLUNTEER WORK
0	26
0	7
15	4
9	1
7	1
0	14
2	6
23	10
7
8

1. Why might the Mann-Whitney U test be preferred to the t test for these data?
2. Use U to test the null hypothesis at the .05 level of significance.
3. Specify the approximate p-value for this test result.