Bunuel wrote:
Scientists have created double-blind studies so that neither the subjects of the experiment nor scientists know whether the subject is receiving an actual drug or a placebo, which is nothing more than a sugar pill. Essentially, if one knows whether one is receiving actual medicine or a placebo, such knowledge can affect the outcome of a study. A recent double-blind study on the effectiveness of the Selective Serotonin Reuptake Inhibitor (SSRI) Freloxamine on depression found that those subjects administered the drug were 15% more likely to have a decrease in symptoms than the control group, which was comprised of those who received a placebo. Since neither group knew which they were receiving, the placebo or the SSRI, the observed drop in depression can only be attributed to Freloxamine.
Which of the following, if true, best calls into question the conclusion of the argument?
(A) Neither the subjects nor the doctors in either group (the control group or the Freloxamine group) knew which drug they were receiving.
(B) Since subjects in both groups were debriefed on the potential side effects of SSRI, which can often be pronounced, many in the Freloxamine group, upon exhibiting side effects, concluded that they were being administered the SSRI.
(C) Freloxamine does not exhibit a uniform effect in all subjects, with many reporting little improvement in symptoms of depression, even after several months of taking the drug.
(D) At dosages two-fold of those employed in the trial, Freloxamine has been shown to cause brief episodes of psychosis.
(E) One subject from the Freloxamine group experienced debilitating side effects and was forced to drop out of the trial before its completion.
Official Explanation:
The credited response is choice (B). The argument states: if one knows whether one has received the actual drug or the sugar pill, that can influence the outcome of the study. The assumption of the experimental design is that no one knows this. But, if folks who get the real drug get side-effects, it won't be hard for them to figure out that they got the real drug, and folks who don't get any side-effects will deduce that they only got a sugar pill. Thus, subjects will know which kind of pill they received, and this knowledge will skew the results of the study.
Choice (A) reiterates the conditions of the design. This conforms to the assumptions of the experiment, so it doesn't call the experiment into question at all. Choice (A) is incorrect.
Choice (C) would account for why some people in the group that received the drug did not see a decrease in symptoms, but the argument does not depend on any kind of claim of universal effectiveness. It merely says that the folks who receive the drug were 15% more likely to show improvement than folks who receive the sugar pill. It is assumed that some people in each group were not helped. Choice (C) does not weaken the argument, so it is incorrect.
What might happen at dosage levels not employed in the experiment is irrelevant to the interpretation of the experiment. Choice (D) is irrelevant, so it is incorrect.
The results of the experiment concerns overall probabilities over a group of people. Scientist don't run studies like this on only, say, six people: this study likely has dozens, if not hundreds, of subjects --- such numbers would be far more typical. If one person had to drop out, that's unfortunate for that person, but that person simply would not be counted: this wouldn’t significantly change the probabilities of the group. Choice (E) is irrelevant and incorrect.
close
15 Nov 2021, 03:30
45%
(medium)
31%
(02:54)
wrong 
