
Arnold F. answered 02/28/16
Tutor
5
(53)
College Professor & Expert Tutor In Statistics and Calculus
If we combine the studies medication A had 306 successes out of 400.
B had 317 successes out of 400.
So if we want to decide which had a better "overall" success rate it would be B.
This seems counter-intuitive with the individual test data and is an example of what is called in statistics "Simpson's Paradox."
Does this make sense?