Most Logical Reasoning questions ask you to do something to an argument: strengthen it, weaken it, find its flaw. Evaluate the Argument questions ask for something stranger. They ask which piece of information would be most useful to have, and the correct answer is a question, not a claim. Because the answer choices are phrased as “whether…” or “how many…,” many test-takers freeze. The usual strategy of finding an answer that supports or attacks the argument does not fit.
The question type is rare, but when it appears it is reliably answerable with one clean test. Once you have it, these questions become some of the most mechanical on the section.
A typical stem reads: “Which of the following would be most useful to know in order to evaluate the argument?” The correct answer identifies a fact that, once known, would tell you whether the argument's reasoning holds up. It is the unknown the whole argument turns on.
That unknown is almost always the argument's assumption, rewritten as a question. If an argument assumes that the people surveyed represent everyone, the useful question is whether the people surveyed were representative. The answer, yes or no, decides how good the argument is.
For each answer choice, imagine the answer coming back as “yes,” and then as “no.” Ask what each outcome does to the conclusion.
| If the answer is… | …and the choice is… | Verdict |
|---|---|---|
| Yes strengthens, no weakens (or the reverse) | Whether the key assumption is true | Correct: the answer decides the argument |
| Yes and no both help the conclusion (or both hurt it) | A one-sided fact that moves the argument only one way | Wrong: the answer does not decide anything |
| Yes and no both leave the conclusion untouched | An interesting but irrelevant detail | Wrong: out of scope |
| Neither answer says anything about this conclusion | A question about a different conclusion | Wrong: wrong target |
Apply the test. (B): if sales rose similarly elsewhere, the lanes probably did little, and the argument is badly weakened. If sales did not rise elsewhere, the argument gets stronger. The two answers point in opposite directions, so (B) is correct. (A) has no bearing on whether sales rose because of the lanes. (C) is the trap: popularity with cyclists sounds relevant, but the answer does not tell you whether sales rose because of the lanes. A “yes” or a “no” leaves the causal question open.
You can predict the shape of the correct answer from the argument type, which saves time because the choices are short but easy to over-read.
| Argument type | The question you should expect |
|---|---|
| Causal (A happened, then B, so A caused B) | Whether B would have happened anyway, or whether something else caused both |
| Survey or sample (some people said X, so everyone feels X) | Whether the sample was representative or self-selected |
| Analogy (this worked there, so it will work here) | Whether the two situations differ in a way that matters |
| Prediction from a plan (the plan will achieve the goal) | Whether the plan can actually be carried out, or has a side effect that undoes it |
| Statistic or comparison (the number rose, so things improved) | Whether the measure means what the author assumes it means |
In each case the prediction is the same move you make on a Necessary Assumption question: find the gap, then phrase it as something unknown. If you have practiced the Denial Test, you already have the skill. Denying an assumption shows what the argument needs; Evaluate asks you to find the question that probes it.
The choice is about something that could only help the argument, however it turns out. “Whether the bike lanes are popular” is a good example. Even a “no” would leave the argument standing. These feel relevant because they touch the topic, but they fail the two-way test.
The choice would be useful for some other conclusion. The argument concludes that lanes boost sales, and the choice asks about the lanes' effect on traffic. Always re-read the conclusion before committing, because Evaluate answers drift toward the topic instead of the claim.
The choice asks something the stimulus already told you. These are tempting because they sound analytical, but they add nothing. A quick scan of the stimulus rules them out.
Take any ten Strengthen or Weaken questions you have already finished. For each one, rewrite the correct answer as a question beginning with “whether.” You will often find that the correct answer converts into an Evaluate-style question with little change. This also trains the underlying skill, which is to separate what an argument claims from what it needs. That separation is useful everywhere in Logical Reasoning. See also our guide to Strengthen and Weaken questions.