Choice B is the best answer. The majority of the passage focuses on the experiment concerning “how much the crowd influences the individual, and whether it can be controlled from outside” (lines 42-44). After explaining the experiment and the results it produced, the passage moves on to consider questions raised by the results, such as whether the findings are site specific or “true in general” (lines 75-76), why different findings are observed, and whether companies can “boost their products by manipulating online ratings on a massive scale” (lines 85-86).

Choice A is incorrect because the passage does not conclude by explaining the practical ways the experiment’s findings have been applied but rather by considering questions the findings raise. Choices C and D are incorrect because the passage does not indicate that there were any flaws in the experiment’s findings and does not include statements from anyone who disputes the findings.

Choice C is the best answer. The author of the passage suggests that a group of people can be “wiser” and more effective than a single person at assessing a quantitative answer, or a measurement, versus producing a valid qualitative judgment, or a judgment of the quality of something. This is most clear in lines 11-14, which state that when guessing a bull’s weight or how many gumballs are in a jar, “your guess is probably going to be far from the mark, whereas the average of many people’s choices is remarkably close to the true number.”

Choices A, B, and D are incorrect because lines 11-14 indicate that the author believes that crowds may be more effective than individuals when arriving at quantitative answers rather than qualitative results. Nothing in the passage suggests that the author believes that crowds are better at starting disagreements than studying an issue in depth (choice A), supporting ideas rather than challenging them (choice B), or ranking opinions rather than coming up with new ideas (choice D).

Choice B is the best answer. The previous question asks what the author of the passage suggests about the wisdom of crowds, with the answer being that crowds can be more effective at producing quantitative answers than qualitative results. This is supported in lines 11-14: when it comes to guessing a bull’s weight or how many gumballs are in a jar, “your guess is probably going to be far from the mark, whereas the average of many people’s choices is remarkably close to the true number.”

Choices A, C, and D are incorrect because the lines cited do not support the answer to the previous question about the author’s belief about when the wisdom of a crowd is effective. Instead, they simply state that crowds are sometimes wiser than individuals, without explaining when (choice A), put forth a theory held by someone other than the author (choice C), and explain how hypotheses about the wisdom of crowds could be tested (choice D).

Choice A is the best answer. In the passage, the author explains that those who are skeptical of the theory that “measuring the aggregate of people’s opinions produces a stable, reliable value” (lines 18- 20) believe that “people’s opinions are easily swayed by those of others” (lines 20-21). This idea is best supported in lines 55-58, which describe a finding from a study of opinions in crowds: “Comments that received fake positive votes from the researchers were 32% more likely to receive more positive votes compared with a control, the team reports.” In other words, people were more likely to give a positive vote when they thought other people had given positive votes.

Choices B, C, and D are incorrect because the lines cited do not provide support for the skeptics’ idea that people’s opinions are easily influenced by the thoughts of others. Instead, they cite findings concerning people giving ratings different from those already given (choices B and C) and share an observation that the degree to which others can be influenced depends in part on the context of the situation (choice D).

Choice B is the best answer. One question Watts asks in regard to the experiment is whether the results would hold true on a larger scale. The passage quotes him in lines 74-76: “‘[But] one question is whether the positive [herding] bias is specific to this site’ or true in general.” Doing the experiment again but collecting ratings on multiple websites would address Watts’s question, as it would show whether or not the same results occur on other sites.

Choices A, C, and D are incorrect. Providing fewer fake positive comments during the experiment (choice A), requiring users to be registered on the website (choice C), or telling users that their answers will be studied (choice D) are actions that likely would affect the results of the experiment involving users voting on comments about stories on one news website, but they would not address Watts’s questions about whether the study would produce the same results on other websites or why different categories of news items had different effects on the news website.

Choice C is the best answer. In lines 85-86 the author asks, “Will companies be able to boost their products by manipulating online ratings on a massive scale?” In the context of selling products by manipulating user reviews, “boost” most nearly means promote. Choices A, B, and D are incorrect because in the context of selling products by manipulating user reviews, the word “boost” refers to promoting the products, not making them larger or bigger (choice A), faster (choice B), or safe (choice D).

Choice A is the best answer. In lines 85-86 the author asks, “Will companies be able to boost their products by manipulating online ratings on a massive scale?” In the context of selling products by manipulating user reviews on a massive scale, the word “scale” most nearly means level or size.

Choices B, C, and D are incorrect because in the context of selling products by manipulating user reviews, a massive “scale” refers to a great level or size, not to a payment (choice B), an interval or space between things (choice C), or a plan (choice D).

Choice B is the best answer. The figure shows that while the mean score of the control comments in the politics category is below 2.0, the artificially up-voted mean score for that category is exactly 2.5. Choice A is incorrect because the artificially up-voted mean score of comments in the business category is higher than 3.0.

Choice C is incorrect because the artificially up-voted mean score of comments in the fun category is less than 2.5. Choice D is incorrect because the artificially up-voted mean score of the comments in the general news category is just over 2.0.

Choice D is the best answer. The figure shows that the mean score for both control comments and artificially up-voted comments in the general news category is just above 2.0. Choice A is incorrect because the mean score for the control comments in the culture and society category is a little below 2.5 while the mean score for the artificially up-voted comments is over 3.0.

Choice B is incorrect because the mean score for the control comments in the information technology category is a little above 1.5 while the mean score for the artificially up-voted comments is above 2.0. Choice C is incorrect because the mean score for the control comments in the fun category is exactly 2.0 while the mean score for the artificially up-voted comments is nearly 2.5.

Choice D is the best answer. In the passage Watts notes that “the category of the news items . . . had a strong effect on how much people could be manipulated” (lines 76-79). That idea is directly supported by the data in the figure, which show that the difference in mean score between the control comments and the artificially up-voted comments varies by subject (for example, in the general news category there is virtually no difference between the mean scores of the two types of comments, while for the business category there is almost a 1.0-point difference between the mean scores).

Choices A and B are incorrect because the passage provides no data for artificially down-voted comments or negative social influence. Choice C is incorrect because the figure applies only to one context (mean score of control comments versus mean score of artificially up-voted comments on the news site); there is no way to tell what patterns would be observed in other contexts.