Problem

In Exercises 6.50 – 6.52, we tested three sets of hypotheses using portions of the same...

In Exercises 6.50 – 6.52, we tested three sets of hypotheses using portions of the same datasets in each of the sets of hypotheses. Let the experiment-wide Type I error rate be defined as the probability of making at least one Type I error in testing any set of hypotheses using the data from the experiment.

a. If we tested each of the three sets of hypotheses at the .05 level, estimate the experiment wide Type I error rate.

b. Suggest a procedure by which we could be ensured that the experiment-wide Type I error rate would be at most .05.

REFERENCE:

6.51 : Refer to Exercise 6.50.

a. Compare the mean drop in blood pressure for the low-dose group and the control group. Use α = .05 and report the level of significance.

b. Estimate the size of the difference in the mean drop for the low-dose and control groups using a 95% confidence interval.

c. Do the conditions required for the statistical techniques used in (a) and (b) appear to be satisfied? Justify your answer.

6.52 : Refer to Exercise 6.50.

a. Compare the mean drop in blood pressure for the low-dose group and the high-dose group. Use α = .05 and report the level of significance.

b. Estimate the size of the difference in the mean drop for the low-dose and high-dose groups using a 95% confidence interval.

c. Do the conditions required for the statistical techniques used in (a) and (b) appear to be satisfied? Justify your answer.

6.50 : A study was conducted to evaluate the effectiveness of an antihypertensive product. Three groups of 20 rats each were randomly selected from a strain of hypertensive rats. The 20 rats in the first group were treated with a low dose of an antihypertensive product, the second group with a higher dose of the same product, and the third group with an inert control. Note that negative values represent increases in blood pressure. The accompanying computer output can be used to answer the following questions.