SAT Command of Evidence: A Repeatable Method to Answer Them Every Time

What "Command of Evidence" Actually Means on the Digital SAT

Command of Evidence is one of the four content domains in the Digital SAT Reading & Writing section, and it shows up more than you might expect — it's tied with Craft and Structure as one of the largest question categories. On a full Reading & Writing section (54 questions across two modules in 64 minutes), you'll see a cluster of these in the Information and Ideas part of each module.

There are exactly two flavors:

• Textual evidence — a short passage makes a claim, poses a hypothesis, or describes a situation, and you pick the sentence that best supports (or, occasionally, weakens) it.
• Quantitative evidence — a passage is paired with a graph, table, or bar chart, and you pick the answer choice whose data correctly completes or supports the argument.

Here's the good news: despite looking different, both types reward the exact same skill — precisely matching an answer to a specific claim. Once you internalize one method, you can answer them consistently and quickly. Let's build that method.

The Core Method: Claim First, Answer Second

The single biggest mistake students make is reading the answer choices before they've locked down what the passage is actually asking them to prove. The choices are designed to sound plausible. If you don't know your target, every option looks tempting.

So use this four-step loop every single time:

1. Find the claim. Locate the exact sentence that states what needs support. It's usually near the end and often starts with a signal like "This suggests…," "Researchers concluded…," "To support this hypothesis…," or "The team predicted that…"
2. Paraphrase the claim in your own words. Compress it to a plain-English version. If you can't restate it, you don't understand it yet.
3. Predict what support would look like. Before touching the choices, ask: "What kind of fact or number would make this claim true?"
4. Match, don't like. Test each choice against your prediction. Eliminate anything that's off-topic, only half-relevant, or contradicts the claim — even if it's a true statement.

That last point is the trap on nearly every question: a choice can be factually accurate and still be wrong because it doesn't support this specific claim.

Worked Example: Textual Evidence

Marine biologist Dr. Favor's team studied whether the cleaner shrimp Lysmata amboinensis recognizes individual client fish rather than treating all fish identically. If the shrimp does distinguish individuals, the researchers reasoned, then it should behave differently toward a familiar client than toward an unfamiliar one of the same species.

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

Run the loop:

• Claim: The shrimp recognizes individual fish.
• Paraphrase: Shrimp treats a known fish differently from a stranger.
• Prediction: I need evidence of different behavior toward familiar vs. unfamiliar individuals.

Now the choices:

• (A) The shrimp cleaned all fish that approached it. → Same behavior for everyone. Contradicts the claim. Eliminate.
• (B) The shrimp began cleaning familiar clients more quickly than unfamiliar ones. → Different behavior toward familiar vs. unfamiliar. Exact match. Keep.
• (C) The shrimp cleaned larger fish more thoroughly than smaller fish. → Size, not familiarity. Off-target. Eliminate.
• (D) The shrimp lived longer when it had access to more client fish. → True-sounding, irrelevant to recognition. Eliminate.

Answer: (B). Notice how the prediction did all the work — you knew the answer's shape before reading a word of the options.

Worked Example: Quantitative Evidence

Quantitative questions look scary because of the graphic, but the method is identical. The added rule: read the axes, labels, and units first, then find the claim, then find the one data point that matches.

A horticulturalist claimed that this seed variety germinates best at moderate soil temperatures, with rates declining at both cooler and hotter extremes. The data ______

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

Run the loop:

• Read the graphic: x-axis is soil temperature (10, 18, 25, 32 °C); y-axis is germination percentage. Values rise to a peak at 25 °C (72%), then drop at 32 °C (40%).
• Claim: Best at moderate temps, worse at both cold and hot extremes.
• Prediction: A choice showing the highest rate in the middle and lower rates at 10 °C and 32 °C.

Choices:

• (A) support the claim, as germination was highest at 25 °C (72%) and lower at both 10 °C (30%) and 32 °C (40%). → Matches shape and cites accurate numbers. Keep.
• (B) support the claim, as germination increased steadily as temperature rose. → False; it drops at 32 °C. Eliminate.
• (C) weaken the claim, as the highest germination occurred at the coldest temperature. → Misreads the graph. Eliminate.
• (D) support the claim, as germination at 18 °C exceeded germination at 25 °C. → 55 < 72, factually wrong. Eliminate.

Answer: (A). The winning choice always does two things: it points in the right direction (support/weaken) and uses numbers you can verify on the graphic.

The Traps, Named So You Can Spot Them

Once you've done a few dozen, you'll notice the wrong answers recycle the same four disguises:

• The True-but-Irrelevant. Accurate, on-topic-ish, but doesn't touch the specific claim. Beaten by a sharp paraphrase.
• The Half-Match. Supports part of the claim but ignores a key condition (e.g., in the shrimp example, a choice about "familiar" fish that never mentions "unfamiliar").
• The Reversal. Correct data, wrong verb — it says "support" when the numbers actually weaken, or vice versa.
• The Misread. In quantitative questions, it swaps axes, misreads a value, or compares the wrong bars. Always check the number against the graphic.

Timing and Habits That Make It Automatic

Command of Evidence questions are among the more time-consuming in Reading & Writing because of the reading load. But you don't have infinite time — roughly a bit over a minute per question on average across the section, and quantitative graphics deserve a few extra seconds up front. Protect that time with these habits:

• Never read choices first. Predict, then look. This alone cuts your per-question time because you stop re-reading tempting distractors.
• For graphics, read every label before the claim. Ten seconds on the axes saves you from the Misread trap.
• Anchor to the exact words. If the claim says "unfamiliar," your answer must engage "unfamiliar." Vague overlap isn't support.
• When two choices remain, ask which one an author could not argue against. The correct answer leaves no gap.

Put It on Repeat

The reason this method works is that Command of Evidence isn't testing outside knowledge — it's testing whether you can hold a precise claim in your head and check candidates against it. That's a trainable reflex, and the fastest way to build it is volume: do sets, notice which trap fooled you, and name it.

If you want that reps-under-real-conditions practice, UnlimitedTests gives you adaptive, Bluebook-style Reading & Writing modules where you can drill Command of Evidence questions back-to-back and see explanations that reinforce the claim-first loop. Run the four steps enough times and it stops being a strategy and starts being instinct — which is exactly what you want on test day.