OFFICIAL EXPLANATION
Exit polls, conducted by an independent organization among voters at five polling locations during a recent election, suggested that the incumbent mayor—a Democrat—was going to lose the election by a wide margin. But, in fact, by the time the final results were tabulated, the incumbent had won the election by a narrow margin.
Which of the following, if true, would explain the apparent contradiction in the results of the exit polls?(A) The people chosen at random to be polled by the independent organization happened to be Democrats.
(B) The exit poll locations chosen by the independent organization were in predominantly Republican districts.
(C) The exit polls were conducted during the afternoon, when most of the districts’ younger voters, who did not support the incumbent mayor, were at work.
(D) The incumbent mayor ran on a platform that promised to lower taxes if elected.
(E) An earlier poll, conducted the week before the election, had predicted that the incumbent mayor would win.
B The key to this statistical argument is to understand that the sampling of the voters might not be representative of ALL the voters. Choice A says the voters chosen by the pollsters happened to be Democrats. But in that case, we would expect that the incumbent mayor, a Democrat, would have been predicted the winner of the election. Choice C says the exit poll was conducted at a time of day in which many people who disliked the mayor could not vote, implying that the actual election results for the incumbent would be worse, or at least no better. Choices D and E are outside the scope of the argument. The best answer is B, which gives a statistical reason for the skewed results: The exit polls were conducted in locations where the incumbent had little support—leaving open the possibility that his results would be stronger elsewhere.