Here is the OE from Kaplan:
When asked to explain a discrepancy in the stimulus, the first step is to locate the discrepancy. The unusual thing in the situation described is that a smaller percentage of the older cars had engine work done. One might think that a greater percentage of older cars would need repairs, but notice that the stimulus refers to which automobiles had repairs, not which needed them. If the cars from the 1960s are probably in more need of repair, why have a greater percentage of 1970s cars received repairs? Maybe you tried to pre-phrase an answer. We’re dealing with percentages again, not raw numbers, so fewer older cars on the road wouldn’t make for a good answer. Maybe you considered the economics associated with repairs: If a car is worth $500, and engine repairs cost $1000, it probably doesn’t make much financial sense to get the car repaired. This is the idea that gets the point: If older cars are not worth the expense, and are scrapped instead of repaired, then the statistics described in the stimulus are understandable. (C) nicely clears up the
discrepancy.
(A) Inspection for emission levels probably would turn up more problems in older cars, which suggests that the older cars would have a greater percentage of repairs, so (A) only deepens the mystery.
(B) New cars are out of the scope based on the dates provided in the passage. Even if we considered the 1970s cars “new” (at least, newer than cars from the 1960s), this choice would, like choice (A), make the facts even less understandable.
(D) Whether or not simplified car engines result in fewer repairs is pretty shaky, but even if they did, (D) would also suggest that the sixties models would need more repairs.
(E) chastises those negligent 60’s car owners for not maintaining their cars, but it doesn’t mention 1970s cars, and thus does nothing to explain the discrepancy.
• When attempting to resolve a discrepancy, be wary of choices that further confound the odd result or finding found in the stimulus.
• When the discrepancy revolves around two groups or things (which is usually the case), quickly eliminate any choice that offers information on only one of them, like (E) here. A true resolution to such a discrepancy must be based on a fact that in some way relates to both groups. _________________