Unformatted text preview: Outcome Disease No Disease Totals Group Placebo S 1 n 1-S 1 n 1 Drug Treatment S 2 n 2-S 2 n 2 If we were constructing a 95% conﬁdence interval for φ = ω 1 /ω 2 , show that the standard error SE(log(ˆ ω 1 / ˆ ω 2 )) = r 1 S 1 + 1 n 1-S 1 + 1 S 2 + 1 n 2-S 2 , is equivalent to the standard error obtained using the formula on page 541. • No problems to do in R (unless you use R for the “pencil and paper” problems) Include all relevant R input and output. Describe the output in your own words. You may ﬁnd the code included in the lab exercises useful....
View Full Document
- Fall '08
- Normal Distribution, Null hypothesis, Input/output