In medical investigations, the ratio h = p₁/p₂ is often of more interest than the difference p₁ − p₂ (e.g., individuals given 

a. Use these facts to obtain a large-sample 95% CI formula for estimating ln(θ), and then a CI for θ itself.

b. The article “Low-Dose Aspirin for Preventing Recurrent Venous Thromboembolism (VT)” (New Engl. J. Med. 2012: 1979–1987) reports a study in which VT recurred in 73 of 411 patients randomly assigned a placebo and in 57 of the 411 assigned to an aspirin regimen. Calculate an interval of plausible values for θ at the 95% confidence level. What does this interval suggest about the efficacy of the aspirin treatment?

Transcribed Image Text:

treatment 1 are how many times as likely to recover as those given treatment 2?). Let 0P₁/P₂. When m and n are both large, the statistic In(0) has approximately a nor- mal distribution with approximate mean value In(0) and approximate standard deviation [(m - x)/(mx) + (n −y)/(ny)]¹/2