5. A 1979 randomized trial was performed comparing a surgical treatment for angi

Question

5. A 1979 randomized trial was performed comparing a surgical treatment for angina pectoris chest pains due to obstruction of the coronary arteries) to a placebo (non-surgical medical management). In the study, 6 of the patients randomized to receive surgery died before they could be operated on. There was a subsequent debate over how best to analyze the data Some favored approach A, in which all patients were analyzed as they were randomized Others favored approach B, in which those 6 patients were excluded on the grounds that they never actually received the surgical treatment that they were randomized to. Which approach do you think is better? Why? 6. True or False: In a hypothesis test, the null hypothesis can be summarized as "nothing is going on besides chance variation 7. The null hypothesis is a hypothesis about (i) the sample (ii) the population 8. Suppose that a scientist carries out 100 hypothesis tests. Unbeknownst to her, in all 100 cases, the null hypothesis is true. If she uses a cutoff of p 0.15 to reject the null, about how many mistakes would you expect that she makes?

Explanation / Answer

5. Approach A is better. Information about the six patients who died will also require analysis to find out whether they had any characteristics (e.g. demographic profile, medical parameters) different from those included in the treatment or the control sample. If they had, the treatment and control samples should be compared on the basis of those parameters to ensure complete randomisation and prevent any bias

6. False. The null hypothesis is that surgery has insignificant effect."nothing is going on" doesnt sound like effect of surgical treatment

7. The null hypothesis is a hypothesis about the population

8. In all 100 cases, null hypothesis is true.

If p < .15 is used to reject it, it means that probability of rejecting it when it is true is 15%.

This means probability of making a mistake is 15%.

Thus, she is expected to make 15% * 100 = 15 mistakes

Related Questions

Navigate

• Browse (All)
• Browse 5
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.