Solving Command of Evidence (Quantitative) Questions on the SAT

What these questions look like on the Digital SAT

This skill sits in the Information and Ideas domain with Central Ideas and Details, Inferences, and its twin, Command of Evidence: Textual. College Board’s framework puts the domain at about 26% of scored Reading and Writing questions (12–14 per test), shared by those skills, with no official count for this one alone.

The official description: students “determine the quantitative evidence (i.e., data from an informational graphic) that best supports a specified claim or point.” The domain description adds that evidence may “support, challenge, or otherwise respond to” a claim, so expect weaken questions too.

Each question pairs a 25–150-word text with a table, bar graph, or line graph, the only graphic types the section uses. The framework says these questions “do not require students to perform calculations”: you find the right data and judge what it shows.

How the questions get harder

College Board’s Skills Insight describes what students in each score band can typically do with graphics:

The method: Claim first, then the table

Don’t browse the figure hoping something jumps out. The question tells you what matters — group, years, direction, comparison. Find only that.

Use Claim → Locate → Compare. Underline the subject, the measure, and the comparison (higher than, declined, largest share). Then find the matching row/column or graph series. Do the comparison last. If a choice cites a true number from the wrong row, it’s still wrong.

A decision method for every stem

1. Read the stem: support, weaken, or complete? That defines “right.”
2. Underline the claim’s parts: group, measure, time, and relationship (higher, rose, varied widely).
3. Turn it into a data test, such as “two cities far apart.”
4. Each choice must be accurate and pass the test.
5. Two survivors? You skipped part of the claim.

How to read tables, bar graphs, and line graphs

Before any number, read the title, the labels, the units, and the legend. Then:

• Tables: each cell is where a row meets a column. Check the caption for units (percent, count, per 1,000) and read any note below; College Board’s sample table warns that rows may not add to 100 because of rounding.
• Bar graphs: bar height is the value. Match the legend to the series, and check where the axis starts: a scale above zero makes small gaps look large.
• Line graphs: the horizontal axis is usually time, and each line is a series. Read direction, peaks, and crossings. A claim that a gap widened means comparing the distance between lines at the start and the end.

Read the right cell: units and categories

Labels lie in wait: “women 18–24” vs. “all women,” “median” vs. “mean,” “percent of trips” vs. “percent of people.” Match units and categories exactly before you celebrate a number that “looks close.” Watch for:

• Count vs. percent: “more people” is a count; “more common” is a share.
• Time frame: a 2015–2020 claim needs those years, not the latest column.

The “complete the claim” question

In College Board’s sample for this skill, the text claims that survey views varied widely by occupation, then ends “for example, ______.” All four choices quote the table accurately, but two cite one occupation and one cites two occupations with the same value. Only one shows wide variation.

Treat the claim as a spec sheet: “varied widely” needs two values far apart. As in Rhetorical Synthesis, several choices are accurate and only one does the job.

Support or weaken a hypothesis

Turn the hypothesis into a prediction: if it’s right, what should the figure show? Support matches the prediction; weaken breaks it. Accurate data that fits the hypothesis can’t weaken it.

Spot the traps in the choices

Classic traps: correct trend for the wrong group; mixing up absolute counts with percentages; saying “most” when the table shows “more than one other category” but not the maximum; and reversing the inequality.

The true-but-irrelevant trap is the big one. Wrong choices are often accurate readings of the figure, as College Board’s sample shows. Ask of every choice: is it accurate, and does it prove this claim? Also:

• Half the claim: the claim has two parts (“rose in A while falling in B”), and the choice proves one.
• Overreach: a cause from a correlation, “all” from a sample, or a forecast from past data.
• Opposite job: data that supports, offered for a weaken question, or the reverse.

Trends vs. single values

“Increased” needs at least two points. “Highest in 2020” needs a comparison across categories or years. Don’t upgrade a single cell into a trend story the figure doesn’t tell. Trend words are exact:

• Increased / decreased: compare the endpoints.
• Steadily / every year: every step goes the same way; one reversal breaks it.
• Peaked: the highest point, with lower values on both sides.
• Gap widened / narrowed: compare the difference at the start and at the end.

Practice until it’s automatic

Practice questions

Here are three Command of Evidence (Quantitative) questions from our problem sets, one at each difficulty level, in the same format as the Digital SAT. Pick an answer to see why it is right or wrong.

Question 1 · Easy

Park rangers use electronic counters to track how many hikers use each trail. In July, rangers compared the counts for four trails in the park. According to the counts, the busiest trail that month was ______

Which choice most effectively uses data from the graph to complete the statement?

Athe lake trail, with 1,840 hikers.Bthe ridge trail, with 2,610 hikers.Cthe canyon trail, with 1,270 hikers.Dthe meadow trail, with 2,150 hikers.

Question 2 · Medium

A microbiologist tested a compound extracted from a common garden plant on cultures of a soil bacterium. She grew one culture with no compound (the control), one with a low concentration of the compound, and one with a high concentration, measuring bacterial density every three hours. She had hypothesized that the compound would slow bacterial growth at both concentrations tested.

Which choice best describes data from the graph that weaken the microbiologist's hypothesis?

AAt 12 hours, the high-concentration culture had a density of about 0.35, far below the control's 1.10.BAt 12 hours, the low-concentration culture had a density of about 1.20, slightly above the control's 1.10.CAt 0 hours, both treated cultures had the same density as the control culture, about 0.05.DThe control culture's density rose from about 0.05 at 0 hours to about 1.10 at 12 hours.

Question 3 · Hard

Agronomist Tomas Hrabal measured soil moisture in plots planted with a rye cover crop and in bare plots over one growing season. Hrabal hypothesizes that the rye helps soil hold water mainly by shading it and slowing evaporation. If that is the main mechanism, the rye plots' moisture advantage over bare plots should be greatest in the driest months, when evaporation losses matter most, and smallest in the wettest months.

Which choice best describes data from the table that would most directly weaken Hrabal's hypothesis?

AIn April, the wettest month, soil moisture in the rye plots exceeded that in the bare plots by only 2 percentage points.BSoil moisture was higher in the rye plots than in the bare plots in every month from April through August.CSoil moisture in both the rye plots and the bare plots was lower in June than in any other month shown.DIn June, the driest month, soil moisture in the rye plots exceeded that in the bare plots by only 1 percentage point.

Quick checks

Shorter drills on the same method.

Frequently asked questions