#confounding

Question 1: Single choice

An orthopedist gives a questionnaire to 30 of his patients who have no joint problems and finds that 20 of them run regularly. He concludes that running lowers the risk of joint problems. What is the main flaw in this reasoning?

He has no comparison of joint problems between runners and non-runners.1
30 patients are too few; the same design with 300 patients would support his conclusion.2
He should have reported the proportion 20/30 with a confidence interval.3
He should have measured running in hours per week instead of yes/no.4

Timed test