Smith et al. reported data on ovarian carcinoma. People had different number of courses of chemotherapy. The five-year survival data for those with 1 to 4 and 10 or more courses of chemotherapy are as follows:

5-year Status

Courses Dead Alive

1 to 4 21 2

10 2 8

(a) Based on this data and using Fisher's exact test, is there a statistically significant association between number of courses of chemotherapy and five-year survival rates at the type I error rate of 0.05? State null and alternative hypotheses and provide a conclusion based on the results.

(b) What is the odds ratio?