Bunuel
A large group of hyperactive children whose regular diets included food containing large amounts of additives was observed by researchers trained to assess the presence or absence of behavior problems. The children were then placed on a low-additive diet for several weeks, after which they were observed again. Originally nearly 60 percent of the children exhibited behavior problems; after the change in diet, only 30 percent did so. On the basis of these data, it can be concluded that food additives can contribute to behavior problems in hyperactive children.
The evidence cited fails to establish the conclusion because
(A) there is no evidence that the reduction in behavior problems was proportionate to the reduction in food-additive intake
(B) there is no way to know what changes would have occurred without the change of diet, since only children who changed to a low-additive diet were studied
(C) exactly how many children exhibited behavior problems after the change in diet cannot be determined, since the size of the group studied is not precisely given
(D) there is no evidence that the behavior of some of the children was unaffected by additives
(E) the evidence is consistent with the claim that some children exhibit more frequent behavior problems after being on the low-additive diet than they had exhibited when first observed
OFFICIAL EXPLANATION
(A) No. The evidence fails to even establish that it was the reduction in food-additive intake which caused the reduction in behavior problems, let alone that it caused a proportionate reduction.
(B) Yes. In order for the conclusion to reasonably follow, we need to know what would have happened had the diet not been changed. Perhaps the same decrease in behavior problems would have occurred even if the diet was not changed. In other words, we need a control group with which to compare the results of the test group.
(C) No. In this situation, percentage is more important than absolute number.
(D) No. This is irrelevant.
(E) No. Since the evidence fails to establish anything, this answer-choice is not incorrect. However, there is a more precise answer-choice.