Command of evidence questions: find the claim first, then make every choice earn it

The command of evidence questions are the ones that take a whole screen. A paragraph about a study, sometimes a table or a graph, then a question that asks which choice would support, weaken, illustrate or best complete something. They are the longest questions on the Reading and Writing side, there are several per module, and in my timing logs they are where a student who is slow on the section is slow. They are also the ones where careful reading pays off most, because the wrong answers are built to reward skimming.

Find the claim before anything else

Every one of these questions is organized around a claim: a hypothesis, a conclusion, an argument someone in the passage is making. The question will name it (“the researchers’ hypothesis”, “the student’s claim”, “Yun’s argument”). Before you read the choices, find that claim in the passage and say it back in one plain sentence. Not the topic of the passage. The specific thing being argued.

“Birds with brighter plumage attract more mates” is a claim. “The passage is about birds” is not. The wrong answers on these questions are almost all about the topic but not about the claim, and the only defense is to have the claim in front of you in your own words.

The two tests every choice must pass

Once the claim is clear, each choice has to pass two tests, and the order matters.

Does it bear on the claim at all? A finding can be interesting, true and from the same study while having nothing to do with the hypothesis. “The researchers also measured the birds’ weight” is a fact. It does not support or weaken anything about plumage and mating. Cross it off.

Does it push in the direction the question asks? If the question says support, the choice has to make the claim more likely. If it says weaken or undermine, the choice has to make it less likely. The longest-standing trap on these questions is the choice that is clearly relevant and clearly about the claim, and points the wrong way. It is the most tempting wrong answer because it passes the first test perfectly.

Do the tests in that order. Relevance first, then direction. Students who start with direction get pulled toward whichever choice sounds most confident.

The three versions

“Which finding, if true, would most directly support the claim?” The choices are hypothetical findings. If true means you do not get to argue with them; you only ask what they would do to the claim. The right answer usually connects the two halves of the claim directly: if the claim is that A causes B, the supporting finding shows A with B, or more A with more B, or no A with no B. The wrong answers show A with something else, or B with a different cause.

“Which quotation most effectively illustrates the claim?” The passage describes a writer’s argument about a text; the choices are lines from the text. The right quotation does what the claim says, rather than mentions it. If the claim is that the poet presents the sea as indifferent to human suffering, the right line shows the sea being indifferent. A line that mentions the sea and a line that mentions suffering are both traps.

“Which choice most effectively uses data from the table to complete the statement?” The quantitative version. Here each choice is a sentence citing numbers, and it has to pass a third test before the other two: is it true according to the table? Read the choice, find the cell it refers to, check the number. At least one choice is usually false, and it is often the one that sounds most like a conclusion. Then, of the true ones, ask which one actually completes the statement in the passage. True but beside the point is the standard wrong answer here.

The traps, named

• Topic match. The choice is about the right subject and the wrong claim. Caught by saying the claim first.
• Wrong direction. Relevant, specific, and it undermines what you were asked to support. Caught by checking direction second, on purpose.
• True but irrelevant. On the table questions, a choice that correctly reads a row that has nothing to do with the sentence being completed. Caught by asking what the sentence needs before reading the numbers.
• Too strong. A choice that proves more than the claim says, or requires something the passage did not establish. The SAT prefers the modest answer that fits exactly over the bold one that overshoots.

The timing

These questions are long and the temptation is to skim the passage and read the choices carefully. Do the opposite. Read the passage once, properly, until the claim is clear, then move through the choices fast, applying the two tests. A command of evidence question done this way takes a minute to a minute and a half, which is on budget. Done choices-first it takes two or three, because every choice sends you back into the passage to check something you would have known if you had read it once.

If your practice test log shows these in the slow column, the fix is not to read faster. It is to read once, for the claim, and let the claim do the work of eliminating. Transitions questions respond to the same discipline, decide first and then look; so do most of the questions on this section.