An analysis of published medical studies about heart attacks (Crossen, 1994, p. 168) noted that in the studies having randomization and strong controls for bias, the new therapy provided improved treatment 9% of the time. In studies without randomization or other controls for bias, the new therapy provided improved treatment 58% of the time.

a. This result suggests it is better not to use randomization in medical studies because it is harder to show that new ideas are beneficial.
b. Some newspaper articles that suggest a particular food, drug, or environmental agent is harmful or beneficial should be viewed skeptically unless we learn more about the statistical design and analysis for the study.
c. This result shows the value of case-control studies over randomized studies.
d. The randomized studies were poorly conducted, or they would have found the new treatment to be better much more than 9% of the time.