PrepTest 155, Section 4, Question 15

What we’re looking for

Working the answers

• A

  We don't know whether the candidate would try to enact reform after being elected. The argument predicts a voting benefit from publicly supporting it. A speculation about whether the candidate would follow through isn't evidence of a mistake in the inference from past voting patterns.

• B

  The observations aren't opposites. One part of the district shows a pattern of voting for the pro-reform candidate; the rest shows no discernible pattern. No pattern isn't a pattern of voting against reform. It also doesn't establish that everyone elsewhere is indifferent. B misdescribes the contrast the candidate reports.

• C

  I can see why we might wonder whether ten years is enough, but the passage doesn't establish that the sample is unusually small. We aren't told how many elections, candidates, or voters are represented. More importantly, even a very large record of this association could leave the candidate's causal inference unproven. Let's look for an answer that identifies that problem directly.

• D
  Correct

  Here is a clearer way to put the central problem: candidates favoring reform received the votes, but did favoring reform cause those votes?

  Imagine the pro-reform candidates always belonged to a party the northeastern voters preferred. Those voters could have chosen them because of their party, not because of their property-tax position. The historical pattern would look exactly as described, yet adopting the tax position alone wouldn't necessarily attract those voters to this candidate. The argument treats the association as evidence that changing this one position will produce the desired voting response.

• E

  The evidence covers elections during the last ten years. It isn't restricted to information from exactly ten years ago. Some of the observations can be recent. We did wonder whether preferences might change, but E gives the wrong account of how old all the evidence is.