How to Master SAT Evidence-Based Reading Questions

SAT evidence questions ask which text or data most directly supports a claim, completes an argument, or justifies a conclusion. The correct answer must satisfy every required element and match the evidence’s scope. A true statement can still be irrelevant.

College Board’s Reading and Writing overview includes Command of Evidence within Information and Ideas.

Use the C-R-E method

1. Claim: rewrite what must be supported.
2. Requirements: split it into necessary parts.
3. Evidence: select the choice that directly establishes all parts.

Do not choose a quotation merely because it repeats topic words.

Textual evidence example

Claim: “The character accepts the offer reluctantly.”

Required elements: acceptance + reluctance. A quotation saying “Mara agreed, but only after a long silence and repeated objections” supplies both. A quotation showing only objections misses acceptance; one showing only agreement misses reluctance.

Claim strength

Evidence that a method may help in one study cannot support “always works for everyone.” Match modal verbs, quantity, population, and conditions.

Underlining some, often, could, only under, and in this sample prevents overstatement.

Quantitative evidence

Read title, axes, units, categories, and legend. Translate the claim into a comparison before looking at choices.

If the claim says Treatment A increased growth more than B, compare change from baseline for A and B—not the highest final value across all groups.

Worked data example

Control growth: 4 cm. A: 10 cm. B: 7 cm. Claim: A produced a larger increase than B. A increased 6 cm; B increased 3 cm. A choice citing only “A reached 10 cm” is relevant but incomplete if the question demands comparison.

Complete the argument

Some questions provide a study and ask for evidence supporting a hypothesis. Identify the predicted result first. If a scientist predicts warmer water speeds development, the strongest evidence compares development time across temperatures while other factors are controlled.

Eliminate five trap types

• Unsupported: plausible but absent.
• Too broad: extends beyond sample/condition.
• Too narrow: supports only part of claim.
• Reversed: shows opposite relationship.
• True but irrelevant: accurate detail that does not prove target.

Use our SAT Command of Evidence guide for more drills.

Paired-text evidence

Map each author’s claim separately. When asked how Text 2 responds to Text 1, choose evidence matching the specific point of agreement or disagreement. Do not infer hostility from a narrow qualification.

A 20-minute practice routine

• 3 minutes: retrieve trap labels;
• 10 minutes: solve six evidence questions;
• 5 minutes: underline decisive words and explain distractors;
• 2 minutes: log one repeated scope error.

After targeted accuracy improves, mix evidence with inference, central idea, and purpose.

Our short-passage strategy guide provides a timed method.

Error-log examples

Weak: “Picked wrong quote.”

Strong: “Claim required both adoption and hesitation; chosen quote showed hesitation only. Next time split claim into requirements.”

Weak: “Graph mistake.”

Strong: “Compared final values instead of change from baseline. Write required comparison before reading choices.”

Timing

Read the question/claim first, then the short passage or data. If two choices remain, point to exact words satisfying each requirement. After about 90 seconds without decisive evidence, choose, flag, and continue.

Use our SAT reading comprehension strategies for broader pacing.

Bottom line

Worked paired-text example

Text 1 argues that a restoration method improves forest diversity. Text 2 agrees diversity rose at one site but notes the study lacked an untreated comparison area. Text 2 would most likely say Text 1’s conclusion is promising but not yet securely causal. An answer saying Text 2 rejects restoration entirely is too strong; one saying both authors draw identical conclusions ignores the methodological criticism.

Break the evidence into two requirements: observed increase + limitation on causal interpretation. This prevents selecting a choice containing only shared topic language.

Build evidence stamina

Mix quotation, graph, table, and paired-text questions. For each, write the claim in six words and the decisive evidence in one line. Speed grows when the proof target is clear, not when students skim faster.

Evidence questions reward precision. Break the claim into requirements, identify the smallest decisive text or comparison, and reject answers that are stronger, weaker, reversed, or merely on topic.

Always preserve the passage’s stated conditions during review.

Prove the answer before eliminating

Elimination is useful, but start by describing the evidence you expect. For a claim about a change over time, predict a comparison between earlier and later values. For a claim about an author's hesitation, predict language that shows both the action and a qualification. This proof target makes traps easier to identify because each choice can be tested against a concrete requirement.

When two answers remain, circle the exact word where their scope differs: all versus some, caused versus associated, the entire population versus the measured sample. Then return to the passage or data and decide which scope is licensed. Do not break a tie using which sentence sounds more academic.

During review, rewrite one tempting wrong answer so it would become supportable. If “the treatment always increases growth” is too broad, a defensible version might be “the treatment increased mean growth in the tested sample.” This exercise exposes the precise overstatement and trains the calibrated language that evidence questions reward.