Assume we are testing at α = 0.05 for all statistical tests
Consider an arthritis pain study. Male and female subjects received an active or placebo treatment for their arthritis pain, and the subsequent improvement (or not) was recorded.
Sex Treatment Improvement
Yes No
Female Active 21 6
Female Placebo 13 19
Male Active 7 7
Male Placebo 1 10
Is there evidence that either sex or treatment is related to pain improvement?
What do the results of the classification table indicate?
What do the results of the Hosmer-Lemeshow test indicate?
Sample Solution