This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: PSTAT 120C: Solutions to Assignment # 3 April 27, 2009 1. (a) In this data we observe that 8 out of the 10 hospitals had higher recovery rates for Drug B rather than for Drug A. Under the null hypothesis that there is no difference between the two drugs, the number of hospitals where drug B has a higher recovery rate should be a binomial with n = 10 and p = 0 . 5. The associated P value for this test P = 2 P { M ≥ 8 } = 2 10 10 + 10 9 + 10 8 2 10 = 0 . 109375 . There doesn’t seem to be a significant difference between the two drugs. (b) The ttest may not be appropriate because the measurement that we have is a percentage and therefore cannot be greater than 100. This sort of restriction can cause the data to be skewed, and have a nonnormal distribution. 2. I want to compare the power of a nonparametric test to a typical normal test. Suppose I have 150 independent observations from a normal distribution with mean μ and standard deviation 10. I want to test H : μ = 0 vs. H A : μ > Let M = the number of observations that are greater than 0, and ¯ x = the average of the 150 observations. (a) The critical value the critical region C 1 = { ¯ x > z * } for α = 0 . 05 can be calculated by figuring P { Z > 1 . 645 } = 0 . 05...
View
Full Document
 Spring '08
 EduardoMontoya
 Statistics, Normal Distribution, Null hypothesis, #

Click to edit the document details