Research reported in the spring of 2002 cast doubt on the effectiveness of arthroscopic knee surgery for patients with arthritis. Patients suffering from arthritis pain who volunteered to participate in the study were randomly divided into groups. One group received arthroscopic knee surgery. The other group underwent placebo surgery during which incisions were made in their knees, but no surgery was actually performed. Follow-up evaluations over a period of 2 years found that differences in the amount of pain relief experienced by the two groups were not statistically significant. (NEJM 347:81 88 July 11, 2002)
a) Why did the researchers feel it was necessary to have some of the patients undergo placebo "surgery"?
b) Because patients had to consent to participate in this experiment, the subjects were essentially selfselected a kind of voluntary response group. Explain why that does not invalidate the findings of the experiment.
c) What does statistically significant mean in this context?