Question

You are interested in a new treatment for inflammation, as measured by C-reactive protein (hs-CRP). You enroll 7 people in your study, then flip a coin for each person. If they get a “heads”, you give them placebo for a month, measure their hs-CRP, wait a second month, and give them your new drug during the third month and measure hs-CRP again. If they get a “tails”, you give them your new drug for a month, measure their hs-CRP, wait a second month, and give them your placebo during the third month and measure hs-CRP again. You are interested in whether your drug has any effect on hs-CRP, on average.

The data are:                          

	hs-CRP on placebo	hs-CRP on new drug
Person 1	3.1	3.1
Person 2	1.2	0.3
Person 3	1.0	0.6
Person 4	1.4	1.3
Person 5	2.3	2.1
Person 6	5.9	5.1
Person 7	0.6	0.3

1. What are the null and alternative hypotheses?
2. In words specific to this context, what would be a type I error? (Please don’t just give me a general definition, I want you to be specific and apply it to this particular study).
3. Conduct a paired t-test and get a p-value. Write a one sentence conclusion based on the results of the test.
4. Conduct a nonparametric alternative to the paired t-test and get a p-value. Write a one sentence conclusion based on the results of the test.

Answer 1