Digital SAT · Guides · November 6, 2026 · 7 min read

SAT Command of Evidence: textual and quantitative questions, tables and graphs, with worked examples

Command of Evidence is the skill in the Reading and Writing section that asks one thing in several forms: which piece of evidence actually backs up (or undercuts) a specific claim. Sometimes the evidence is a finding from a study, sometimes a line from a poem or novel, and sometimes a number from a table or graph. The method is the same in every case: pin down the claim first, decide what the evidence would have to show, and only then look at the options.

Where it sits in the test

Command of Evidence belongs to the Information and Ideas domain, about 26% of Reading and Writing, together with Central Ideas and Details and Inferences. Information and Ideas questions come after Craft and Structure in each module, so you usually meet Command of Evidence in the first half of a 27-question module, roughly between questions 8 and 16. Expect several per module, split between the two kinds:

For how this domain fits with the other three, see the SAT Reading and Writing guide.

Textual evidence: a four-step method

1. Find the exact claim. Read the question first, then the text. Underline (or note) the precise claim: who does what, under which condition. “Urban birds sing higher because of traffic noise” is a different claim from “urban birds sing higher”.
2. Predict what the evidence must show. Ask: if this claim is true, what would we expect to observe? For a cause, the best evidence usually shows that the effect changes when the cause changes. For a comparison, it shows both sides of the comparison.
3. Treat every option as true. “If true” means you do not judge whether a finding is realistic. You judge only whether it bears on the claim.
4. Match the strength and the scope. The right answer touches every key part of the claim. An option that is about the same topic but leaves out the key condition is wrong, however interesting it is.

For weaken questions, flip step 2: what observation would be hard to explain if the claim were true? Often it is the effect appearing without the cause, or the cause present without the effect.

For illustrate questions with a quotation from a poem, play or novel, break the claim into its parts. If the claim says a speaker “admires the sea’s power but fears it”, the right line shows both admiration and fear. A line with only one of the two is the classic half-right trap.

Quantitative evidence: how to read a table or graph fast

Data questions look heavier than they are. The text tells you which part of the graphic matters; most of the numbers are there to distract you.

• Read the claim before the graphic, then the title, axis labels, units and legend.
• Find the comparison the claim needs. Most claims compare: two groups, two years, a trend over time, or a relationship between two variables. Locate only those cells or bars.
• Check each option on two tests. First, is it accurate according to the data? Second, does it support this claim? Many wrong answers pass the first test and fail the second.
• Watch the units and the time frame. “Percent of households” is not “number of households”; “2015” is not “2025”. Options often swap a column or a year.

The traps that make wrong answers tempting

• True but irrelevant. The option correctly reports the data or a real-sounding finding, but it does not address the claim. This is the most common trap in quantitative questions.
• Half the claim. It supports one part of a two-part claim: the change in water use, but not the link to the fixtures; admiration, but not fear.
• Wrong comparison. It compares the right numbers across the wrong groups, or compares the right groups in the wrong year.
• Opposite direction. It would weaken when you need to support, or the reverse. Reread the question word before choosing.
• Too broad. It is about the topic in general, not the specific relationship in the claim.

Worked example 1: textual evidence

This question was written by Nomto; it is not from an official test.

Many songbirds that live in cities sing at a higher minimum pitch than members of the same species living in the countryside. A team of researchers hypothesizes that urban birds raise their pitch because low-frequency traffic noise would otherwise drown out the lower notes of their songs.

Which finding, if true, would most directly support the researchers’ hypothesis?

A) Urban songbirds sing more often in the early morning than rural songbirds of the same species do.

B) Within one city, birds near busy roads sang at a higher minimum pitch than birds in quiet parks, and the gap narrowed on days with little traffic.

C) Rural songbirds of the same species have larger song repertoires than urban songbirds do.

D) The average level of traffic noise in many large cities has increased over the past decade.

The claim: traffic noise causes the higher pitch. What the evidence must show: pitch goes up where noise is high and comes down when noise drops.

Answer: B. It ties pitch to traffic in two ways: across places (roads versus parks) and across time (busy versus quiet days). A and C describe differences between urban and rural birds but say nothing about noise or pitch. D is the too-broad trap: it is about traffic noise, but not about birds at all.

Worked example 2: quantitative evidence

Average daily water use per household (liters)

Town	2015	2025	Households with low-flow fixtures, 2025
Arden	410	330	72%
Bexley	395	380	18%
Corlis	420	345	65%
Dunmore	380	372	22%

A council member argues that installing low-flow fixtures in homes was associated with a reduction in household water use between 2015 and 2025.

Which choice most effectively uses data from the table to support the council member’s claim?

A) In 2015, households in Arden used more water per day on average than households in Dunmore did.

B) Arden and Corlis, where most households had low-flow fixtures, cut average daily use by 80 and 75 liters, while Bexley and Dunmore, where about a fifth did, cut it by only 15 and 8 liters.

C) Of the four towns, Corlis had the highest average daily water use per household in 2015.

D) In 2025, households in Bexley used more water per day on average than households in any other town.

The claim links two things: the share of homes with fixtures and the drop in use. What the evidence must show: towns with more fixtures saw bigger drops.

Answer: B. Check the numbers: 410 − 330 = 80, 420 − 345 = 75, 395 − 380 = 15, 380 − 372 = 8. High-fixture towns cut use far more than low-fixture towns, which is exactly the association claimed. Now notice that A, C and D are all accurate: 410 is more than 380, 420 is the highest 2015 figure, and 380 is the highest 2025 figure. Each is true but irrelevant, because none of them mentions the fixtures or the change over time. That is why “is it true?” is never enough on its own.

Timing

Reading and Writing gives you about 71 seconds per question on average. Textual Command of Evidence questions usually fit in that time. Data questions can take 80–100 seconds, so win the time back on quick questions such as Boundaries.

• 0–20 seconds: read the question and find the claim. For data, read the title, labels and units.
• 20–50 seconds: predict what the evidence must show; check options against accuracy and relevance.
• If two options survive: choose the one that covers every part of the claim. Stuck after about 1.5 minutes? Pick one, mark it for review and move on; there is no penalty for a wrong answer.

How to practice

Train the two kinds separately until the method is automatic: textual evidence practice and quantitative evidence practice. After each question, name the trap behind the option that tempted you. When your accuracy is steady, mix the skills with the whole Information and Ideas set and then try a timed module in Nomto’s practice test. Its scores are estimates of your level, not official results, so finish with the College Board’s full-length practice tests in Bluebook before test day.