George Petersen gives us a little lesson on how not to use statistics. He knows that 9 out of 10 Da Volos are brought to his shop because of malfunctioning windows, so he reasons that this particular Da Volo, which is being towed in, has probably also been brought in for malfunctioning windows. Surely the fact that the car is being towed indicates that there must be some more serious problem. Petersen has mindlessly applied a numerical formula while ignoring additional information. Where else do we see such reasoning?
(B) uses previous figures to conclude that there's a 9 to 1 chance that Nash's history of World War II is a work of fiction. (B) ignores the compelling contrary evidence (that this book is a history) and mindlessly applies a numerical formula where it clearly shouldn't be applied.
(A)'s use of statistics is reasonable. We don't know anything special about Janine Davis; she's just a voter. Therefore, since almost 3 out of 4 voters chose Lieberman, there's an almost 3 out of 4 (or 3 to 1) chance that Janine voted for Lieberman.
(C) reasons that Vladimir's chances of being admitted into the academy have been improved by the rejection of the previous candidates. That's not a persuasive line of argumentation, but it's nothing like the stimulus.
(D) is a straight numerical argument. The conclusion seems overstated (even a 1 out of 50 chance isn't "virtually no chance"), but it's not at all like the stimulus; we're not shown a particular case with special information that goes against the numbers.
(E)'s mistake is to assume that all those who took more than four years to graduate did not go on to further schooling; this allows (E) to conclude that the 95% who graduated in four years did go on. It's statistically flawed, certainly, but no special case is discussed, and therefore there's no similarity to the stimulus.
