Tricky! The key to this one is to keep track of the argument.
A) Irrelevant. Even if they did not appreciate the ban, it does not tell us anything about the authors conclusion of "smokers having to abandon non-smokers contributed to a decrease in smoking".
B) Trap. The argument is concerned with smoking. Even if the consumption of other forms of tobacco increased, smoking may or may not decrease.
C) Hard to eliminated. However, for C to work we have to make a lot of assumptions - we must assume that the sole cause of the decrease in second hand smoke lung cancer is due to the lack of public smoking. However, there are infinite possible reasons as to why the number may have decreased. Out.
D) Almost. For this work, we have to assume that non smoking teenagers, that smoking teenagers hang out together, that the separation of the two groups induced the smoking group to stop smoking, and that the teenagers composed a majority of those numbers. Too many assumptions are required for D to work
E) The last choice. It does work. The author assumes that smokers, if they abandon non smoking friends, are inclined to stop smoking so as to not abandon their friends. If it is proven that non smoking friends accompany the smokers, then the authors argument becomes more plausible. If the opposite, weaken.
E must be the answer.
Posted from my mobile device