A student reads a short passage about a historian’s argument, glances at four answer choices, and picks the one that sounds the most like the passage. The choice is on topic. It uses the same vocabulary. It even repeats a phrase the author used. And it is wrong, because it supports a claim the question never asked about. That single habit, choosing the option that matches the subject rather than the option that matches the exact claim, costs more command of evidence points than any reading-comprehension weakness ever will.

This guide treats command of evidence as one skill wearing two costumes. The first costume is textual: you are handed a claim and asked which quotation, finding, or statement best supports or undercuts it. The second is quantitative: you are handed the same kind of claim and a table or graph, and asked which data point does the supporting or the undercutting. The verbal surface looks different. The underlying move is identical. Pin the claim down to its exact variable, group, and direction, then test every option against that pinned claim rather than against the general topic. The InsightCrunch two-step method that this article builds, define the claim precisely, then evaluate only for direct support, turns a question type that feels like a judgment call into something closer to arithmetic. Precision about the claim is the whole game, and precision can be drilled.

By the end you will have a reusable screening checklist that runs the same way on a quotation as it does on a bar chart, a set of fully worked items across both subtypes, and a named taxonomy of the traps the writers reuse: the right-topic-wrong-variable trap, the right-topic-wrong-group trap, and the wrong-direction trap that flips support into contradiction. None of that requires you to know the history, the science, or the social science a given passage happens to discuss. Everything you need sits on the screen in front of you. The challenge is reading it with discipline.

Where command of evidence sits on the digital exam

Command of evidence belongs to the Information and Ideas content area of the Reading and Writing section, the same family that holds central-ideas questions and inference questions. The College Board groups it there for a reason. All three ask you to reason from what a text or a dataset actually says, not from what you happen to believe about the world. Where central-ideas questions ask what the passage is mainly doing and inference questions ask what must follow from it, command of evidence asks a narrower and more testable thing: given this specific claim, which piece of information moves it, and in which direction.

The digital format splits the skill into two labeled subtypes. The textual variant gives you a short passage, names a claim or a research goal, and offers four candidate statements, any of which might appear in or extend the text. Your job is to find the one that most directly supports, illustrates, or weakens the named claim. The quantitative variant gives you a passage paired with a graphic, usually a table, a bar chart, or a line graph, and asks which data point best completes, supports, or undercuts the claim the passage sets up. The College Board’s own framework files the quantitative items under the same skill code as the textual ones, which is the clearest signal that they expect you to use one method for both.

Is command of evidence in the Reading section or the Writing section?

It sits in the Reading and Writing section, specifically in the reading-focused Information and Ideas group rather than the grammar-focused Standard English Conventions group. There is no separate reading test and writing test on the digital exam; everything is one Reading and Writing section delivered across two modules. Command of evidence questions can appear in either module.

Both subtypes are reading questions in spirit even when a graphic is involved, because the quantitative item does not ask you to compute anything beyond comparing values. You will never need to run a regression or calculate a standard deviation to answer one. You read the table the way you read a sentence, looking for the one number that bears on the one claim, and you reject the numbers that bear on a neighboring claim instead. That is why the test-design literature describes the quantitative variant as reading-comprehension-heavy work dressed in numbers rather than as a math task. The arithmetic involved rarely goes beyond comparing two figures or recognizing whether a value rose or fell.

The skill also threads through the rest of the section. When you read a science passage and must connect a finding to a hypothesis, you are doing command of evidence even if the question label says something else, which is why the approach here pairs naturally with the method for connecting on-page data to a claim in science passages. When you read a passage with an embedded table, the reading-and-data integration that tables and graphs demand is the same matching discipline applied to a longer text. Treat command of evidence not as an isolated question type but as the spine of evidence-based reading across the whole section.

How often does command of evidence appear?

Command of evidence is one of the more frequently sampled Information and Ideas skills, and both the textual and quantitative variants show up on essentially every administration. Rather than chase an exact tally, which the College Board does not publish and which would be misleading to state as a fixed number, treat it as a high-yield skill worth dedicated practice. Because the same method answers both subtypes, every hour you invest pays off twice.

The frequency matters for pacing. If a skill recurs across the section and resolves to one repeatable procedure, the student who has automated that procedure banks points quickly and saves time for the genuinely interpretive questions where judgment is unavoidable. The student who treats each evidence question as a fresh puzzle burns time and invites the topic trap. The point of this article is to make the procedure automatic so that command of evidence becomes a place where you gain time rather than lose it.

The mechanics up close: what the question actually asks

Strip away the passage subject and a command of evidence item reduces to three parts. There is a claim, stated in the passage or in the question stem. There is a direction, support or weaken, named in the stem. And there are four candidates, each a sentence or a data point that relates to the topic in some way. The correct candidate is the one that bears on the claim itself, in the direction the stem requests, more directly than any other. Everything else in the item is engineered distraction.

The claim is where almost all the difficulty lives, because the claim is usually more specific than students register on a first read. Consider a stem built around the assertion that a particular species of frog called more frequently after rainfall than during dry spells. A careless reader collapses that to “the passage is about frog calls and rain” and then accepts any choice that mentions frogs, calls, or rain. But the claim has parts. It names a frequency comparison, more often versus less often. It names two conditions, after rainfall and during dry spells. It names a single species. A choice that reports how loudly the frogs called, not how frequently, fails the variable. A choice that compares two different species fails the group. A choice that says calling decreased after rain fails the direction. Only the choice that hits the frequency, the two named conditions, and the rising direction can be correct.

What does “best supports the claim” actually require?

It requires a direct, on-target relationship, not merely a true or related statement. A choice can be factually accurate, drawn straight from the passage, and still wrong, because supporting a claim means providing information that makes that specific claim more likely to hold, not information that happens to live nearby. The word “best” signals that more than one choice may be loosely relevant; you want the tightest fit.

This is the hinge of the whole question type, so it is worth stating as a rule. Relevance to the topic is not the standard. Relevance to the claim is the standard. The writers build wrong answers by manufacturing topical relevance, that warm feeling of recognition you get when a choice repeats the passage’s keywords. They count on you mistaking that recognition for support. The defense is to refuse to evaluate any choice until you have written the claim down in your own words, reduced to its variable, its group, and its direction. Once the claim is pinned, the topical decoys lose their power, because you are no longer asking “does this sound related” but “does this move this exact assertion in this exact direction.”

The quantitative variant works the same way, with a graphic standing in for the four sentences. The passage sets up a claim, often by introducing a researcher and the question they investigated, and the graphic holds the numbers. The stem asks which value, comparison, or trend from the graphic supports or weakens the claim. The decoys are real numbers from the same table or chart that describe a neighboring quantity: the wrong row, the wrong column, the wrong year, the percentage when the claim is about a raw count, or the largest value when the claim is about a change over time. The reading discipline is to pin the claim’s exact quantity first, then locate only that quantity in the graphic, and only then read the answer choices.

Why is an on-topic answer often wrong?

Because the test writers deliberately build distractors that match the passage’s subject while missing its precise claim. An on-topic choice triggers recognition, and recognition feels like correctness, so students select it without checking whether it supports the actual assertion. The fix is to evaluate against the pinned claim, never against the general topic.

Reading the structure this way also explains why command of evidence rewards slowing down at exactly one moment and speeding up everywhere else. The expensive part is parsing the claim; the cheap part is screening the choices once the claim is clear. Students who reverse that, skimming the claim and agonizing over the choices, get the worst of both. They spend their time where the test wants them to waste it and skip the step that would have made the choices sort themselves. The mechanics, in short, are an invitation to invest thirty seconds in the claim so that the remaining choices take ten seconds each.

Where does the claim live, the passage or the stem?

It can live in either, and noticing which changes how you pin it. Sometimes the passage states the assertion outright, often near its end, as a researcher’s conclusion or an author’s position, and the stem simply asks which choice supports or weakens that stated assertion. Sometimes the stem builds the assertion itself, handing you a hypothesis, a prediction, or a position that the passage set up but did not fully state, with phrasing like which finding would support the idea that, or which result would be expected if. In the second case the assertion you must pin is in the question, not the text, and reading only the passage leaves you guessing.

The practical habit is to read the stem with the same care you give the passage, because the stem frequently carries the exact assertion and its qualifiers. A stem that asks which result would be expected if a material conducts heat better at higher pressures has handed you a causal, conditional assertion to pin: variable, heat conduction; condition, higher pressure; direction, better. The passage supplies context and the candidate information, but the stem supplies the target. Students who skim stems to get to the choices faster routinely pin the wrong assertion and then wonder why no choice fits cleanly.

A related case appears in the science-style textual items, where the passage presents a hypothesis and asks which experimental result would support it. There the assertion is the hypothesis, and supporting it means predicting the result the hypothesis implies before reading the choices. This is the same prediction discipline the self-coaching drill builds, applied at the moment of reading: decide what a confirming result would look like, then find the choice that reports it. Whether the assertion sits in the passage or the stem, the four-part pin is identical; only the place you read it from moves.

The core investigation: the two-step method for both subtypes

Here is the method this article is built to teach, stated plainly and then demonstrated until it is second nature. Step one: define the claim precisely, reducing it to the smallest set of testable parts, its variable, its group or subject, its direction, and any time or condition the claim attaches to. Step two: evaluate each choice for direct support or direct weakening of that pinned claim, in the stem’s requested direction, discarding any choice that drifts to a different variable, a different group, a different time, or the wrong direction, no matter how topical it sounds. That is the InsightCrunch two-step claim-precision method, and it runs identically on a quotation and on a chart.

Step one in depth: pinning the claim

A claim has anatomy. The variable is the thing being measured or asserted: a frequency, a quantity, a preference, a cause, a magnitude, a presence or absence. The group is who or what the claim is about: a species, a population, a sample, a region, a historical figure, a class of materials. The direction is the shape of the assertion: more or less, increased or decreased, supported or refuted, present or absent. The condition is any qualifier the claim leans on: a time period, a treatment, a comparison group, a circumstance. When you can state all four out loud in a single breath, the claim is pinned.

Pinning forces you to notice the qualifiers that wrong answers feed on. A passage might claim that a new fertilizer raised crop yield in the first growing season but not in later ones. A reader who pins that claim notices the condition, first season only, and is immediately armed against any choice describing yields in the third season, no matter how impressive those yields are. A passage might claim that a poet’s late work grew more pessimistic in tone than the early work. The pinned reader notices the comparison, late versus early, and rejects any choice that quotes a gloomy line from the middle period, because the middle is neither term of the comparison. The discipline is almost mechanical, and that is the point: a mechanical first step protects you from the intuition the test is trying to exploit.

Step two in depth: screening the choices

Once the claim is pinned, run every choice through a fixed screen rather than reading them as a menu of plausible options. Ask of each choice in turn, does this concern the claim’s variable, does it concern the claim’s group, does it sit in the claim’s condition or time, and does it push in the direction the stem requested. A choice that fails any of those four checks is out, and you do not need to deliberate further about how nicely it is written or how much it echoes the passage. A choice that passes all four is your answer, and if two choices both seem to pass, you have not pinned the claim finely enough and should return to step one to find the distinction the writers built in.

This screen is the findable artifact of the article, and it is worth carrying as a small mental table. The same four columns, variable, group, condition, direction, apply whether the choices are sentences or data points. Below is the screen laid out as a checklist you can rehearse until the columns run automatically.

Screen check	Question to ask the choice	Eliminates this trap
Variable	Does this address the exact quantity or quality the claim is about, not a neighboring one?	Right-topic-wrong-variable
Group	Does this describe the same subject, population, or sample the claim names?	Right-topic-wrong-group
Condition	Does this fall within the claim’s stated time, treatment, or comparison?	Right-topic-wrong-time
Direction	Does this push the claim the way the stem asks, support when it says support, weaken when it says weaken?	Wrong-direction reversal

The screen is deliberately blunt. Its power comes from being applied before you let yourself like or dislike a choice. Most wrong selections happen because a student reads the choices first, forms a favorite based on how it sounds, and then looks for reasons to keep it. The screen reverses that order. You decide what the answer must do before you read what the choices say, so the choices walk into a test they did not get to design.

How the weaken variant changes the screen

The weaken variant changes only the direction column, and it is worth treating carefully because students who have drilled support sometimes run the support screen out of habit. When the stem asks which finding would weaken or undermine the claim, the correct choice still must concern the right variable, group, and condition; it simply must push against the claim instead of for it. A weaken item about the frog claim, that calling rose after rainfall, would be answered by the choice showing calling did not rise, or fell, after rainfall in that species. A choice showing some unrelated decline, or a decline in a different species, fails the group or the variable even though it points downward. Direction alone is never enough; the downward push has to land on the claim.

The most common weaken-variant error is selecting a choice that is merely neutral or irrelevant, on the theory that anything failing to support the claim must weaken it. Neutrality is not weakening. A choice that says nothing about the claim’s variable neither props it up nor knocks it down; it sits outside the question entirely. To weaken, a choice must engage the claim and then contradict it. Hold that distinction and the weaken variant becomes the support variant with one sign flipped, which is exactly how the College Board intends it to behave.

Why one method spans two subtypes

The reason the textual and quantitative variants yield to a single method is that the test writers are testing the same cognitive act in both: the disciplined matching of a specific assertion to specific information. A sentence and a data point are both just carriers of information. The frog claim could be supported by a sentence reporting more frequent calls after rain, or by a table row showing a higher call count in the rainy condition than the dry one. The reasoning that selects the right sentence is the reasoning that selects the right row. Students who internalize this stop treating the graphic as a separate beast requiring math nerves and start treating it as a sentence written in numbers.

This unity is the article’s namable claim, the thing other study pages can cite: command of evidence is a single skill, claim-precision, expressed across two surfaces, and the right-topic-wrong-specifics family of traps is the same family in both. Once you believe that, your practice consolidates. You are not learning two techniques and hoping the right one fires under pressure. You are learning one technique and applying it to whichever surface the screen happens to show you. That consolidation is where the time savings and the accuracy gains both come from.

The four claim shapes and how each is supported

Pinning a claim becomes faster once you recognize that almost every assertion the section builds an evidence question around falls into one of four recurring shapes. Knowing the shape tells you in advance what supporting information has to look like, so you screen the choices against a template rather than from scratch. This is a practical extension of the two-step method: the shape sharpens step one and pre-loads step two.

Causal claims

A causal assertion says one thing produced, caused, raised, lowered, or led to another. The fertilizer raised yield; the program reduced trips; wave action promoted growth. Supporting a causal assertion means showing the effect tracked the cause: when the cause was present the effect appeared, and ideally when the cause was absent the effect did not. The strongest supporting choice ties the outcome to the proposed cause specifically, while the most dangerous decoy introduces a different possible cause for the same outcome, a confounder. When a stem asks you to weaken a causal assertion, the cleanest weakening choice either shows the effect occurred without the cause or shows the cause occurred without the effect. Treat causal assertions with extra care for the condition column, because causation almost always carries a comparison, with the cause against without it, and the writers love to answer it with data from only one side of that comparison.

Comparative claims

A comparative assertion ranks or contrasts two or more things: more than, faster than, higher than, the most, the least. The post-illness symphonies used smaller orchestras than the earlier ones; banking trust exceeded the other sectors. Supporting a comparison requires evidence that names both terms, or at least establishes the side the assertion says won. The signature decoy here cites a true value for one term while staying silent on the other, which feels supportive but proves nothing about the contrast. A second decoy compares the wrong pair, swapping in a term the assertion never mentioned. When you pin a comparative assertion, write down both terms explicitly, because the comparison is only as supported as its weaker half.

Trend claims

A trend assertion describes change over an ordered sequence, usually time: rose steadily, declined, peaked, leveled off, grew across all four quarters. Supporting a trend requires evidence consistent with the whole shape the assertion names, not just one favorable point. This is why a claim of steady increase is weakened by any single interruption, and why a claim of decline is weakened by any genuine rise within the span. Trend assertions live almost entirely in the quantitative variant, paired with line graphs, and the reading skill is to trace the entire described span rather than anchor on the endpoints. A line that starts low and ends high can still dip in the middle, and a trend assertion of uniformity cannot survive that dip.

Existence and magnitude claims

The fourth shape simply asserts that something is present, absent, large, small, or notable: a measurable difference exists, an effect was substantial, a category was rare. Supporting these requires evidence that the named quantity reached the asserted level, and weakening them requires showing it did not. The decoy of choice here is the magnitude swap, a choice that confirms something exists but at the wrong size, or that reports a different quantity at the asserted size. Existence and magnitude assertions reward attention to the exact adjective: substantial, slight, negligible, dramatic, because the supporting evidence must match the strength of the word, not merely its direction.

Claim shape	What support must show	Signature decoy
Causal	Effect tracked the proposed cause, present and absent	A different cause for the same effect
Comparative	Both terms, or the winning side, by the stated measure	One term cited, the other left silent
Trend	The full described shape across the ordered span	One favorable endpoint, ignoring an interruption
Existence or magnitude	The named quantity reached the asserted level	The right thing at the wrong size

Recognizing the shape is a speed tool. The instant you classify an assertion as causal, you know to watch for a confounder; the instant you classify it as a trend, you know one contrary segment can decide a weaken item. The template does part of step two before you have read a single choice.

Worked examples: the textual variant

The examples that follow use invented passages and claims so the mechanics stay visible. Real test items wrap the same structure in denser prose, but the bones are identical. For each item the claim is pinned first, then the choices are screened, and the principle that generalizes is named at the end. Reading these slowly, with the screen in hand, is worth more than skimming a dozen.

A textual support item

The passage reports that a marine biologist, studying a reef off the coast, found that a particular coral species grew faster in waters with moderate wave action than in calm, sheltered lagoons. The stem asks which finding, if true, would most directly support the biologist’s conclusion that moderate wave action promotes growth in this coral.

Pin the claim. Variable: growth rate. Group: this one coral species. Condition: moderate wave action compared against calm water. Direction: support, meaning growth should be higher under waves.

Now screen. A choice reporting that the coral’s color was brighter in wave-exposed zones fails the variable; color is not growth. A choice reporting that a different coral species grew faster in calm water fails the group and the direction both. A choice reporting that fish were more abundant near the wave-exposed coral fails the variable entirely; abundance of fish says nothing about coral growth. The choice reporting that colonies in the wave-exposed zone added more skeletal mass per month than colonies in the sheltered lagoon passes every column: it concerns growth, the right species, the wave-versus-calm condition, and it points upward under waves. That is the answer. The generalizable principle: a supporting choice must restate the claim’s own variable in the claim’s own direction, not a pleasant correlate of it.

A textual weaken item

The same passage, but now the stem asks which finding would most directly weaken the conclusion that moderate wave action promotes growth in this coral.

The claim is pinned identically, growth, this species, waves versus calm, but the direction now flips: the correct choice must push growth down under waves or up under calm, contradicting the conclusion.

Screen the choices. A choice that the wave-exposed colonies and the sheltered colonies showed no difference in skeletal mass directly contradicts the claim, because it removes the very advantage the conclusion asserts; it concerns growth, the right species, the right comparison, and it pushes against the claim. That is the weakening answer. A choice that the wave-exposed colonies were more colorful is irrelevant, not weakening. A choice that a related species grew slower in calm water actually leans toward supporting the general idea and is the wrong direction besides. A choice that the lagoon water was warmer introduces a new variable without addressing growth’s relationship to wave action; it is a tempting confounder but it neither supports nor weakens the stated claim on its own. The principle: weakening requires engaging the claim’s variable and then contradicting it, and a confounding detail that changes the subject is not the same as a finding that opposes the claim.

A right-topic-wrong-variable trap

A passage describes a study of urban trees, claiming that streets lined with mature trees recorded lower midday air temperatures than treeless streets. The stem asks for the finding that best supports the claim that mature street trees lower midday air temperature.

Pin it. Variable: midday air temperature. Group: streets with mature trees versus treeless streets. Direction: support, temperature lower with trees.

The seductive wrong choice reports that tree-lined streets had higher property values than treeless ones. It is deeply on topic, trees and streets and a measurable benefit, and a tired reader nods along. But property value is not air temperature; it fails the variable. Another wrong choice reports that tree-lined streets had more pedestrian traffic at midday; again on topic, again the wrong variable. The correct choice reports that thermometers on tree-lined streets read several degrees lower at noon than thermometers on treeless streets. It alone restates the claim’s variable, temperature, in the claim’s direction. The principle this item teaches is the article’s central warning: the most dangerous distractor is not the obviously irrelevant one but the one that swaps the variable while keeping the topic, because topic familiarity masquerades as support.

A right-topic-wrong-group trap

A passage about language learning claims that adult learners who practiced speaking daily improved their pronunciation faster than adult learners who only studied grammar. The stem asks which result best supports the claim that daily speaking practice accelerates pronunciation improvement in adult learners.

Pin it. Variable: rate of pronunciation improvement. Group: adult learners, speaking-practice group versus grammar-only group. Direction: faster for the speaking group.

The trap choice reports that child learners who practiced speaking daily reached native-like pronunciation within a year. It is glowing, it is on topic, and it concerns speaking practice and pronunciation, but it describes children, not adults. It fails the group. The claim is bounded to adult learners, and evidence about children, however striking, does not bear on it. The correct choice reports that among the adults, the daily-speaking group showed measurably faster pronunciation gains than the grammar-only group over the same period. It hits the variable, the bounded group, and the direction. The principle: a claim’s group is a wall, and evidence from the other side of the wall is not evidence for the claim, regardless of how favorable it appears.

A claim-precision item with a hidden qualifier

A passage claims that a composer’s symphonies written after a serious illness used smaller orchestras than the symphonies written before it. The stem asks which observation best supports the claim that the post-illness symphonies favored smaller orchestras.

Pin it carefully, because the qualifier is the trap. Variable: orchestra size. Group: this composer’s symphonies. Condition: after the illness compared with before. Direction: smaller after.

A choice reporting that the composer’s post-illness chamber pieces used only a handful of players is on topic and points to smallness, but chamber pieces are not symphonies; the claim is bounded to symphonies, so this fails the group within the group. A choice reporting that one early symphony used an unusually large orchestra is true and relevant to the comparison but, standing alone, it does not establish the post-illness shrinkage; it describes one term of the comparison without the other. The strongest choice reports that the post-illness symphonies called for noticeably fewer instrumentalists than the pre-illness symphonies, naming both terms of the comparison in the claim’s direction. The principle: when a claim hangs on a comparison, the supporting evidence must address both terms or at least the side the claim asserts changed, and a true fact about an adjacent category does not satisfy it.

A paired comparison of two near-identical choices

The hardest textual items leave two choices that both seem to pass a loose screen, and the writers count on you flipping a coin. A passage claims that a city’s bike-share program reduced short car trips downtown. Two choices survive a careless screen. The first reports that downtown car trips under two miles fell after the program launched. The second reports that overall downtown traffic congestion eased after the program launched. Both are on topic, both point downward, both occur after launch.

The distinction lives in the variable. The claim is specifically about short car trips, trips you might otherwise have biked, not about congestion in general. Congestion can ease for many reasons unrelated to short trips, a new transit line, a changed work-from-home pattern, a construction project ending. The first choice restates the claim’s exact variable, short car trips, while the second substitutes a broader, looser variable that the claim did not assert. The first choice wins. The principle, and the resolution rule for every near-tie: when two choices survive, the correct one matches the claim’s variable more exactly, and the loser almost always broadens or narrows that variable just slightly. Tightness, not loudness, decides.

A causal textual item with a confounder decoy

A passage reports that a school added a free breakfast program and that attendance rose the following term, leading the principal to conclude the breakfast program improved attendance. The stem asks which finding, if true, would most strongly support the conclusion that the breakfast program improved attendance.

Pin it as a causal assertion. Variable: attendance. Group: this school’s students. Condition: with the program compared against without it. Direction: the program raised attendance. Because it is causal, the supporting evidence should tie the rise specifically to the breakfast program and not to a coincident change.

Screen the choices. A choice reporting that the school also hired three new teachers that term is a textbook confounder; it offers a rival cause for the same attendance rise and therefore does nothing to support the breakfast link, and it arguably undercuts it. A choice reporting that students liked the breakfast food is about preference, not attendance, and fails the variable. A choice reporting that attendance at a neighboring school without a program stayed flat over the same term is the quiet winner: by showing the effect did not appear where the cause was absent, it strengthens the case that the program, not some district-wide factor, drove the change. The principle: causal assertions are best supported by evidence that isolates the proposed cause, and the with-versus-without comparison is worth more than any single glowing number about the treated group alone.

A causal weaken item using the same passage

Keep the breakfast passage and ask instead which finding would most weaken the conclusion that the program improved attendance.

The pin is identical; the direction flips. To weaken a causal assertion, show the effect without the cause or the cause without the effect, or supply a rival cause that better explains the outcome. The strongest weakening choice reports that a nearby school with no breakfast program saw the same attendance rise that term, which points to a shared external cause and strips the program of its explanatory role. A choice reporting that a few students skipped breakfast is too small to undo the conclusion and does not address attendance directly. A choice reporting that the program was expensive concerns cost, not effect, and fails the variable. The principle: the cleanest way to weaken a cause-and-effect assertion is to show the effect occurring where the cause is missing, because that single contrast dissolves the link the conclusion depends on.

Worked examples: the quantitative variant

The quantitative items replace the four sentences with a graphic, but the screen does not change. You still pin the claim’s variable, group, condition, and direction, and you still reject data that describes a neighboring quantity. The only added skill is reading the graphic accurately enough to locate the right cell, bar, or point, which is itself a discipline of checking labels before values. Practicing these against an unlimited bank of fresh sets, the kind of section-targeted, immediately graded practice the SAT Reading and Writing tool at ReportMedic provides, is the fastest way to make label-checking reflexive.

Reading the graphic before the claim tempts you

Before any quantitative item, read the graphic’s furniture, not its numbers. Read the title, which names what the whole display measures. Read the column headers and the row labels in a table, or the axis labels and units in a chart. Read any note beneath the graphic, because notes carry the qualifiers that decide close items, a rounding caveat, a sample size, a unit conversion. Only after the furniture is clear should you let yourself read individual values, and only after the claim is pinned should you decide which value matters. Students who read numbers first end up anchored to whatever figure caught their eye, and the writers place an eye-catching figure in the wrong row on purpose.

A quantitative support item from a table

A passage introduces researchers surveying public trust in automated tools across four service sectors, and presents a table of trust percentages by sector. The passage builds toward a claim that trust was highest in the banking sector. The stem asks which choice, using the table, best supports that claim.

Pin it. Variable: trust percentage. Group: the four sectors, with banking as the focus. Condition: the surveyed population. Direction: banking should be the highest.

Read the furniture: the table’s rows are sectors, the single data column is percent expressing trust, and a note warns that rows may not total a round figure because of rounding. Now screen the choices, each of which cites a number from the table. A choice citing the retail sector’s percentage fails the group focus; retail is not banking. A choice citing banking’s percentage but describing it as the lowest fails the direction. A choice that adds two sectors’ percentages together computes a quantity the claim never mentioned and fails the variable as the claim framed it. The correct choice cites banking’s percentage and identifies it as the largest of the four, which is exactly the claim restated with the supporting value attached. The principle: in a table item, the right answer names the claim’s cell and characterizes it the way the claim does, and the decoys cite real cells that answer a different question.

A bar-chart comparison item

A passage describes a survey of how four age brackets prefer to receive news, and presents a bar chart with one cluster of bars per bracket. The passage builds toward the assertion that the youngest bracket relied on mobile devices more than any other bracket. The stem asks which feature of the chart supports that assertion.

Pin it as a comparative claim. Variable: share relying on mobile devices. Group: the four age brackets, youngest as the focus. Direction: youngest highest on mobile.

Read the furniture: the horizontal axis groups the brackets, the vertical axis is percent relying on a given device, and each bracket has separate bars for mobile, desktop, and print. The comparison the assertion makes is across brackets within the mobile bars only. The decoy invites you to read the tallest bar anywhere in the chart, which might be an older bracket’s desktop bar, and call it support. That fails the variable, because it leaves the mobile category the assertion is about. Another decoy reports the youngest bracket’s desktop share, right group, wrong device. The supporting feature is that the youngest bracket’s mobile bar stands taller than every other bracket’s mobile bar. The principle: in a clustered bar chart, a comparative claim restricts you to one category of bars across groups, and the visually dominant bar is frequently planted in the category the claim does not concern.

A line-graph inflection item

A passage claims that a lake’s measured oxygen level fell sharply only after the water temperature crossed a particular threshold, and pairs the claim with a line graph plotting oxygen against rising temperature. The stem asks which feature of the graph supports the claim that the sharp decline began only past that threshold.

Pin it as a trend claim with a condition. Variable: oxygen level. Group: this lake’s measurements. Condition: temperatures above the named threshold. Direction: sharp fall only after the threshold, implying little change before it.

Read the furniture: the horizontal axis is temperature, the vertical axis is oxygen, and the plotted line shows the relationship. The supporting feature is the inflection point, the place where the line, roughly level or gently sloped before the threshold, bends into a steep downward run after it. A decoy reports that oxygen was highest at the lowest temperature, true but silent on where the sharp fall begins. Another decoy reports the single lowest oxygen value, which sits at the far end and says nothing about onset. The principle: trend claims that name a turning point are supported by the shape of the line at that point, the inflection, not by its maximum or minimum, and reading a line graph for command of evidence means reading where the slope changes, not just where the line is tallest or shortest.

A quantitative weaken item from a graph

A passage claims that a tutoring program raised test scores steadily across all four quarters of a school year, and pairs the claim with a line graph of average scores by quarter. The stem asks which feature of the graph, if accurate, would most weaken the claim of steady improvement across all four quarters.

Pin it. Variable: average score. Group: the program’s students. Condition: across all four quarters. Direction: the claim says steadily rising, so weakening means showing the rise was not steady or not present in some quarter.

Read the furniture: the horizontal axis is quarter one through quarter four, the vertical axis is average score, and each plotted point is one quarter’s mean. Now screen. A choice noting that scores rose from quarter one to quarter two supports the claim, wrong direction for a weaken item. A choice noting that scores were highest in quarter four is also support. A choice noting that a different school’s scores fell fails the group. The weakening choice notes that the average score dropped between quarter two and quarter three before rising again, because a mid-year drop contradicts the assertion of steady improvement across all four quarters. It engages the exact variable, the exact group, the full condition, and pushes against the claim. The principle: a weaken-from-graph item is answered by the point or segment that breaks the pattern the claim asserts, and a single contrary segment is enough to undo a claim of uniformity.

A percentage-versus-raw-number trap

A passage claims that more people visited the city’s central library than its largest branch library last year, and presents a table showing, for each location, the raw number of visits and the percentage change from the prior year. The stem asks which value best supports the claim that the central library drew more visitors than the largest branch.

Pin it. Variable: number of visitors, a raw count. Group: central library versus largest branch. Direction: central higher.

The trap is built from the percentage column. A choice citing that the largest branch posted a higher percentage growth than the central library is true from the table and feels like it favors the branch, but percentage growth is not visitor count; a small branch can post large percentage growth on a small base. It fails the variable. The correct choice cites the raw visit counts and shows the central library’s count exceeded the branch’s count. The principle, and one of the most reliable traps in the quantitative variant: when a claim concerns a raw quantity, percentage figures in the same table are decoys, and when a claim concerns a rate or change, raw counts are the decoys. Match the claim’s unit before you match its number.

A wrong-row, wrong-column misread caught in time

A passage claims that, among three crops studied, the one with the highest water requirement also had the highest yield per acre, and presents a table whose rows are crops and whose columns are water requirement and yield per acre. The stem asks which table value supports the linkage the claim draws between high water use and high yield.

Pin it. Variable: the pairing of water requirement and yield for the same crop. Group: the three crops. Direction: the highest-water crop should also top the yield column.

The misread the writers invite is to grab the highest number in the yield column and the highest number in the water column without checking that they belong to the same row. If the crop with the highest water requirement is not the crop with the highest yield, the claim fails, and the supporting answer must be a choice that identifies a single crop topping both columns. The defense is to read across the row, not down the column: find the crop that holds the top water value, then look across its own row to confirm it also holds the top yield value, and select the choice that names that crop on both counts. The principle: table claims that link two columns are answered row-first, never column-first, because the writers separate the two top values into different rows to catch the column-skimmer.

A “which statement is accurate based on the table” reading item

Some quantitative items do not ask for support or weakening at all; they ask which statement is accurate according to the graphic. These are pure reading items, and the screen simplifies to two columns, does the statement describe the right cell, and does it describe that cell correctly. A passage presents a table of rainfall by month for two cities, and the stem asks which statement is accurate based on the table. A choice claiming city A had more rain than city B in every month is accurate only if the table shows that across all twelve rows; one exception falsifies it. A choice claiming city A’s wettest month exceeded city B’s wettest month requires comparing each city’s own maximum, not the same calendar month. The correct choice is the one whose every clause survives a literal check against the cells. The principle: accuracy items reward suspicion of universal words, every, always, never, highest, because a single contrary cell breaks a universal claim, and the writers love a sweeping statement that one row quietly refutes.

A pie-chart proportion item

A passage reports how a museum’s annual budget divided across five categories and claims that conservation consumed a larger share than education. The accompanying pie chart shows the five slices. The stem asks which feature of the chart supports the claim that conservation took a larger share than education.

Pin it as a comparative claim about proportion. Variable: share of the budget, a proportion. Group: the conservation and education categories. Direction: conservation slice larger than education slice.

Read the furniture: a pie chart encodes proportion by slice size, so the comparison is purely between two named slices. The decoy points at the largest slice overall, which may be neither conservation nor education, and calls it support; that fails the group, since the claim names two specific categories, not the budget leader. Another decoy cites a raw dollar figure if the chart includes one, drifting from proportion to amount when the claim is about share. The supporting feature is simply that the conservation slice is visibly larger than the education slice. The principle: pie charts answer proportion claims by direct slice comparison between the two named categories, and the dominant slice is a distraction unless the claim happens to name it.

One claim, two surfaces: the parallel demonstration

To make the unity of the method concrete, here is a single assertion run through both subtypes side by side. A passage about commuting reports a researcher’s finding that workers who switched to a four-day week reported less commuting stress than those who kept a five-day week. The assertion to support: the four-day schedule lowered commuting stress relative to the five-day schedule.

Pin it once, and the pin serves both surfaces. Variable: reported commuting stress. Group: workers, four-day group versus five-day group. Condition: after the schedule change. Direction: lower stress for the four-day group.

In the textual version, the four candidates are sentences. The supporting sentence reports that the four-day group recorded lower average stress scores than the five-day group after the switch. A decoy reports that the four-day group enjoyed their weekends more, a different variable. A decoy reports that a separate study of students found less stress with fewer class days, a different group. A decoy reports that the four-day group’s stress rose, the wrong direction. Only the first sentence restates the pinned assertion in full.

In the quantitative version, the four candidates point at a table of average stress scores by group and week. The supporting choice cites the cell showing the four-day group’s average stress below the five-day group’s average. The decoys cite the four-day group’s productivity figure if present (wrong variable), a different occupation’s row (wrong group), or a figure showing the four-day group higher (wrong direction). The correct data point and the correct sentence are mirror images: both carry the same comparison, in the same direction, for the same group, on the same variable.

That parallel is the heart of the article’s claim. You did not switch methods between the two versions. You pinned once and screened twice. The sentence and the table cell were interchangeable carriers of a single piece of supporting information, and the screen sorted them identically. Internalize this and the quantitative variant stops being a separate, scarier question type and becomes the textual variant with numbers in the candidate slots.

The trap taxonomy: how wrong answers are built

Naming a trap is half of defeating it, because a named trap is something you can scan for deliberately rather than fall into blindly. The command of evidence distractors come from a small, reusable family, and once you can label them on sight you stop deliberating and start eliminating. This is the InsightCrunch right-topic-wrong-specifics taxonomy, and it covers nearly every wrong answer the section produces in this question type.

The right-topic-wrong-variable trap

This is the workhorse distractor and the one that catches the most students. The choice stays squarely on the passage’s subject but swaps the measured quantity for a neighbor: temperature becomes property value, growth becomes color, frequency becomes loudness, count becomes percentage. Because the topic is intact, the choice feels relevant, and relevance is the feeling the trap is engineered to produce. The defense is the variable column of the screen: name the claim’s exact quantity before reading the choices, and reject any choice that measures something else, however interesting. A useful habit is to underline the claim’s variable in your own restatement, so that when a choice offers a different noun you feel the mismatch immediately.

The right-topic-wrong-group trap

Here the choice keeps the variable and the direction but shifts the subject: adults become children, this species becomes a related species, the central library becomes a branch, the post-illness symphonies become the chamber works. The evidence is often genuinely strong for the group it describes, which is what makes it tempting; strength is not relevance. The defense is the group column: a claim’s subject is a boundary, and evidence from outside the boundary does not count, no matter how vivid. When a passage carefully specifies a population, treat that specification as load-bearing, because the writers specified it precisely so they could build a wrong answer about the population next door.

The right-topic-wrong-time or wrong-condition trap

This trap exploits the qualifier. A claim bounded to the first growing season is answered wrongly with data from the third season; a claim about midday temperature is answered wrongly with morning data; a claim about behavior after rainfall is answered wrongly with behavior during a drought. The choice matches the variable and the group but slips the condition. The defense is the condition column, and the underlying habit is to treat every temporal or circumstantial qualifier in the claim as a tripwire. If the claim says “after,” any “before” or “during” choice is suspect on sight, and you can often eliminate it before fully reading it.

The wrong-direction reversal

The reversal keeps the variable, the group, and the condition but points the wrong way: on a support item it offers a choice that would weaken the claim, or on a weaken item a choice that would support it. Under time pressure, with the right topic and the right numbers in view, students grab a correct-feeling choice without checking which way it pushes. The defense is the direction column, and the simplest guard is to say the stem’s direction word aloud, support or weaken, immediately before screening, so the requested direction stays loaded in working memory while you read.

The neutral-but-irrelevant decoy

Subtler than the others, this choice neither supports nor weakens; it simply sits beside the claim saying something true and unrelated. On weaken items especially, students reason that a choice failing to support must therefore weaken, and select a neutral statement. It does neither. The defense is to insist that a correct choice must engage the claim’s variable and then take a side; a statement that takes no side on the variable is outside the question. When a choice leaves you thinking “that is true, but so what,” trust the “so what,” because the writers planted it to absorb a guess.

A graded view of the family

Trap name	What stays right	What goes wrong	The screen column that catches it
Wrong-variable	Topic, group, often direction	Swaps the measured quantity	Variable
Wrong-group	Variable, direction	Shifts the subject or population	Group
Wrong-time or condition	Variable, group	Slips the qualifier or comparison	Condition
Direction reversal	Variable, group, condition	Pushes against the stem’s request	Direction
Neutral decoy	Topic relevance	Takes no side on the variable	Any; the choice fails engagement

Carrying this table in your head turns elimination into a checklist rather than a feeling. When you face four choices, you are no longer asking which one you like; you are asking which traps the writers built this time, and that question almost answers itself once the family is familiar.

Pacing: where to spend your seconds

Command of evidence rewards an unusual time distribution, and getting that distribution right is worth as much as knowing the method. The expensive moment is pinning the claim, and the cheap moments are everything after. A reader who pins well spends a deliberate beat parsing the assertion into its four parts and then moves through the choices quickly, because pinned claims make most distractors fall on contact. A reader who skips the pin saves a few seconds at the start and then loses far more thrashing between choices that all sound plausible. The arithmetic of the section punishes the false economy of a rushed claim.

Practically, aim to read the passage once at a steady pace, stopping the instant you reach the stated claim or the question stem that names it. Restate that assertion to yourself in one sentence with its variable, group, condition, and direction before your eyes touch the choices. For the quantitative variant, insert one extra step between reading the passage and reading the choices: read the graphic’s labels and note. Then run the screen. If a choice survives all four columns, mark it and verify quickly that no other choice survives too; if two survive, you under-pinned and should return to the assertion for the distinction you missed rather than rereading the choices a third time.

How long should a command of evidence question take?

There is no official per-question budget, since the digital section gives you a total time for the module rather than a stopwatch on each item, but a useful working target is to keep evidence questions at or slightly under your average pace so they bank time for the genuinely interpretive items. Pinning the claim should take a handful of seconds; screening four choices should take only a little longer once the assertion is fixed. The questions feel slow only when the pin is skipped.

The deeper pacing lesson is that command of evidence is one of the few places on the section where method directly buys speed. On a tone question or a hard inference, even a strong method leaves real interpretive work that takes time. On an evidence question, a clean pin collapses the choices almost mechanically, so the better your method the faster you go, with no accuracy cost. That is why this question type belongs near the front of any time-management plan: it is where disciplined readers make up the seconds they will need elsewhere. Build the pinning habit through repetition until it runs without conscious effort, and the pacing takes care of itself.

How command of evidence differs from inference and central ideas

Because all three question types live in the Information and Ideas group and all three reward close reading, students blur them, and the blur causes errors. Each rewards a different relationship to the text, and knowing which relationship a question wants keeps you from importing the wrong instinct.

Central-ideas questions ask what the passage is mainly about or what its primary point is. They reward stepping back to the whole and rejecting choices that fixate on a single detail. The danger there is the too-narrow choice, a true detail mistaken for the main point. Inference questions ask what must be true given the passage, what follows even though the text does not state it outright. They reward a strict must-it-follow test and reject choices that are merely plausible or reasonable in the real world. The danger there is the overreach, a likely-sounding statement the passage does not actually license.

Command of evidence sits between these. Unlike central ideas, it does not want the gist; it wants the one piece of information that bears on a specified assertion. Unlike inference, it does not ask what follows from the passage; it asks what would support or weaken a claim the question hands you, often a claim you evaluate against information you have not yet been given. The instinct command of evidence rewards is matching, not summarizing and not deducing. You are not asked what the passage means overall, and you are not asked what you can conclude from it; you are asked which candidate does the most for a named assertion in a named direction.

Why does mixing these instincts cause errors?

Mixing them causes errors because each wrong instinct selects a different wrong answer. Bring the central-ideas instinct to an evidence question and you reach for the broadest, most summary-like choice, which usually fails the variable by being too general. Bring the inference instinct and you reach for the most reasonable real-world statement, which often is not the one that most directly supports the specific claim. The evidence question wants neither breadth nor plausibility; it wants targeted, direct relevance to the pinned assertion.

Holding the three apart is itself a skill worth drilling, and it pays beyond accuracy. When you can name what a question is asking for, you stop running a generic close-reading routine on every item and start running the specific routine each type rewards. For command of evidence that routine is the two-step method, and the method only fires correctly when you have first recognized that you are in an evidence question and not a central-ideas or inference question wearing similar clothes. Practice sets that mix the three types, with the labels hidden, are the best way to train that recognition, because they force you to diagnose the question before answering it, exactly as the live section will.

Common mistakes and the misconception that drives them

The single misconception underneath nearly every command of evidence error is the belief that the question rewards topical matching. Students read the passage, absorb its subject, and then hunt for the choice that most resembles what they just read. That instinct works on weaker reading questions and fails here, because the writers have learned to make the most resemblant choice the wrong one. The correction is to replace topical matching with claim matching as a conscious, named substitution. You are not looking for the choice that sounds most like the passage; you are looking for the choice that does the most for the specific claim.

A second mistake is reading the answer choices before pinning the claim. The moment you read four plausible options, your mind forms a favorite, and everything after that is rationalization. The order matters enormously: claim first, choices second, every time. A third mistake is treating the quantitative variant as a math problem and freezing. There is no math to fear; comparing two numbers or reading a trend is the whole arithmetic, and the difficulty is entirely in matching the right number to the claim. A fourth mistake is ignoring the graphic’s note, that small line of text beneath a table or chart that often holds the qualifier deciding a close item. A fifth is rushing the universal words in accuracy items, accepting an “in every month” or “always” claim without checking each relevant cell.

What is the most common command of evidence mistake overall?

The most common single mistake is selecting an on-topic choice that addresses the wrong variable, because it feels supported when it is merely related. Students recognize the passage’s keywords in the choice and stop checking. The cure is to pin the claim’s exact variable before reading any choice and to reject anything that measures a neighboring quantity.

Underneath all five mistakes runs the same root, which is impatience with step one. Pinning the claim takes a deliberate beat, and under a ticking clock the deliberate beat feels like a luxury you cannot afford. The data says the opposite: the beat spent pinning is repaid several times over in the choices that then eliminate themselves, while the beat skipped is paid for in agonized coin-flips between near-identical options. The student who learns to spend time on the claim and almost none on liking or disliking choices is the student who turns this question type into reliable points.

A self-coaching drill to make the method automatic

Method fails under pressure when it has not become a habit, and habits form through narrow, repeated reps rather than through reading about them. Here is a drill that builds the pinning reflex in short sessions, using nothing but practice passages you already have. Take any passage with an evidence question and cover the answer choices with your hand or a card. Read only the passage and the stem. Out loud or on scratch paper, state the assertion in one sentence and tag its four parts: variable, group, condition, direction. Predict, before uncovering anything, what the correct choice has to do. Only then reveal the choices and find the one matching your prediction.

The prediction step is the engine of the drill. When you force yourself to say what support would look like before seeing the options, you build an expectation that the decoys then have to beat, which they usually cannot. Students who skip prediction let the choices set the terms; students who predict make the choices audition for a role already written. After each item, do not just check right or wrong. Name which trap the wrong choices were, using the taxonomy: this one swapped the variable, that one shifted the group, the third reversed direction. Labeling misses converts a vague sense of difficulty into a specific, fixable pattern, and patterns are what improvement is made of.

Run the drill in both subtypes deliberately. Alternate a textual item with a quantitative one so your brain learns that the pin does not change when the surface does. For quantitative reps, add the label-reading step: before predicting, read the graphic’s title, headers, units, and note aloud, so that habit also automates. Keep the sessions short, ten or fifteen items, and frequent, because the reflex you are building is fast retrieval under time pressure, and fast retrieval comes from many small bouts rather than a few long ones.

How do I know the method has become automatic?

You will know the method has set in when you find yourself pinning the claim before you have consciously decided to, when the four parts surface without effort and the obviously off-variable choices fall away before you have finished reading them. At that point the question type stops feeling like reading and starts feeling like sorting, which is the sensation you are training toward.

A second sign is your error pattern changing. Early on, misses come from grabbing on-topic decoys; with the method ingrained, the rare miss tends to come from a genuinely close pair where the writers buried a subtle variable distinction, which is a far better problem to have and a far rarer one. Track your misses by trap label across a few sessions, and watch the wrong-variable and wrong-group columns shrink. When they do, you have moved the skill from something you do deliberately to something you do reflexively, and that shift is the entire return on the drill.

What to study first, and the verdict on practice

If you have limited preparation time and want the fastest return on command of evidence, the verdict is unambiguous: drill step one in isolation before you touch full items. Take any passage, find a claim or build one, and practice stating its variable, group, condition, and direction in a single sentence, over and over, until pinning is reflexive. This is the bottleneck skill, and strengthening it improves not only command of evidence but inference questions and central-ideas questions too, because all three reward the same precise reading. The technique transfers directly to the must-be-true discipline behind inference and implication questions and to the topic-versus-argument distinction at the heart of main-idea and purpose questions, so the hour you spend here pays dividends across the Information and Ideas family.

Once pinning is solid, drill the two subtypes together rather than separately, because the whole point is that one method serves both, and alternating textual and quantitative items trains your brain to apply the screen regardless of surface. Mix in a deliberate diet of weaken items, since most students over-practice support and then misfire on direction under pressure. And practice with timing only after the screen is automatic; speed built on a shaky method just produces faster mistakes. The right sequence is accuracy first through the screen, then speed through repetition, the same order the strategy for rhetorical synthesis questions recommends for that adjacent skill and the same order any serious plan to score 1500 or higher follows across the section.

For the practice itself, you want volume and immediate feedback, because the screen only sets in through repetition with correction. Working through a large, section-targeted set of textual and quantitative items with full worked solutions, then reviewing every miss against the trap taxonomy to label which column failed you, is the routine that moves the skill. The SAT Reading and Writing practice tool at ReportMedic is built for exactly that rhythm, giving you fresh evidence items with answer explanations so you can turn each reading into rehearsal and each miss into a named, fixable pattern rather than a vague sense of having gotten it wrong.

The larger verdict is this. Command of evidence looks like a reading question and behaves like a logic puzzle with a fixed solution procedure. Students who keep treating it as fuzzy reading keep scoring it fuzzily. Students who adopt the two-step method, pin the claim, then screen the choices, and who learn to name the trap family on sight, convert one of the more frequent question types on the section into a dependable source of points. The method is not clever and it is not secret. It is just disciplined, and discipline is the rarest thing in a timed test and therefore the most valuable.

Carry one sentence out of this article and let it run on every evidence item you ever face: do not answer the topic, answer the claim. The passage will always offer you a choice that matches the subject and misses the assertion, and the day you stop reaching for that choice is the day this question type starts paying you back. Pin the variable, the group, the condition, and the direction; screen what is in front of you against those four; and trust the screen over the warm pull of a familiar word. The students who win these points are not better readers in some mysterious sense. They are simply readers who decided what the answer had to do before they let the choices tell them what to think.

Frequently asked questions

What is a command of evidence question on the SAT?

A command of evidence question gives you a claim and asks which piece of information best supports it, illustrates it, or weakens it. It lives in the Information and Ideas group of the Reading and Writing section and comes in two labeled forms. The textual form offers four sentences or findings and asks which one most directly bears on the named claim. The quantitative form pairs the passage with a table or graph and asks which data point does the supporting or weakening. In both, the correct choice is the one that addresses the claim’s exact variable, group, condition, and direction, not merely the passage’s general topic. The skill being tested is precise matching of a specific assertion to specific evidence, which is why an on-topic but off-claim choice is the classic wrong answer. Treat it as a logic task with a fixed procedure rather than as open-ended reading, and it becomes one of the most predictable question types on the section.

What is the difference between textual and quantitative evidence?

The difference is only the surface that carries the evidence. Textual evidence questions hand you four sentences, quotations, or research findings and ask which best supports or weakens the claim. Quantitative evidence questions hand you a table, bar chart, or line graph and ask which value, comparison, or trend does the same job. The reasoning is identical in both: pin the claim to its variable, group, condition, and direction, then select the piece of information that matches all four. People often fear the quantitative form because it shows numbers, but it is not a math question. The arithmetic never exceeds comparing two values or noticing whether something rose or fell. The genuine work is reading the graphic’s labels carefully and matching the right cell or point to the claim, which is the same discipline you use to match the right sentence in the textual form. Because one method answers both, you should practice them together rather than as two separate skills.

How do I find the quote that best supports a claim?

Start by ignoring the choices and pinning the claim. State its variable, the thing being measured or asserted, its group, the subject it applies to, its condition, any time or comparison it leans on, and its direction, which way it points. Hold that pinned claim in mind. Then read each candidate quote and ask whether it addresses the same variable, concerns the same group, sits within the same condition, and pushes in the direction the stem requested. The quote that passes all four checks is the answer. Reject any quote that swaps the variable, even slightly, or that describes a neighboring group, or that drifts to a different time, no matter how much it echoes the passage’s wording. The most common error is choosing the quote that sounds most like the passage rather than the one that does the most for the specific claim. Resemblance is the trap; targeted relevance is the standard.

What is the two-step method for command of evidence?

The two-step method is the core technique this guide teaches. Step one is to define the claim precisely by reducing it to its smallest testable parts: the variable being asserted, the group it applies to, the condition or time that bounds it, and the direction it points. You state these in a single sentence before you read any answer choice. Step two is to evaluate each choice strictly against that pinned claim, asking whether it matches the variable, the group, and the condition, and whether it pushes in the stem’s requested direction. A choice that fails any check is eliminated without further deliberation; a choice that passes all of them is the answer. The method works identically for textual and quantitative items, because a sentence and a data point are just two carriers of the same kind of information. Its power comes from the order: deciding what the answer must do before reading what the choices say, so the distractors cannot seduce you with topical familiarity.

How do I answer a “which data weakens the claim” question?

Pin the claim exactly as you would for a support item, naming its variable, group, and condition, but set your target direction to opposition rather than agreement. The correct choice must engage the claim’s own variable and group and then contradict the claim’s assertion. If the claim says a treatment raised a value, the weakening data shows the value did not rise, or fell, for that exact group under that exact condition. Two traps are common here. The first is selecting a choice that points downward but describes a different group or variable, which fails the match even though it feels contrary. The second, more frequent, is selecting a neutral choice on the theory that anything not supporting the claim must weaken it; neutrality is not weakening. To weaken, a choice has to take a side against the specific assertion. Say the word “weaken” to yourself before screening so the direction stays loaded, then demand both a match and a contradiction from the choice you pick.

Why is an on-topic answer sometimes wrong here?

Because the test writers build wrong answers to be on topic on purpose. A choice that repeats the passage’s keywords triggers a feeling of recognition, and recognition feels like correctness, so students select it without checking whether it actually supports the claim. The defense is to separate two ideas that the trap deliberately blurs: relevance to the topic and relevance to the claim. The topic is the broad subject the passage discusses; the claim is the narrow assertion the question asks about. A choice can be entirely on topic, even drawn straight from the passage, while addressing a different variable, group, or time than the claim names, which makes it useless as support. Pin the claim in your own words first, then judge each choice against that pinned claim rather than against the general subject. Once you do, the on-topic decoys lose their pull, because you are asking whether the choice moves this exact assertion, not whether it sounds related.

How do I read a graph for a quantitative evidence question?

Read the graph’s furniture before its figures. Start with the title, which tells you what the whole display measures. Then read the axis labels and units on a chart, or the row and column headers in a table, so you know what each position represents. Read any note beneath the graphic, because notes carry qualifiers, a rounding caveat, a sample size, a unit, that decide close items. Only after the structure is clear should you read individual values, and only after you have pinned the claim should you decide which value matters. For tables, read across the relevant row rather than down a column when a claim links two quantities, because the writers separate the two highest values into different rows to catch column-skimmers. For line graphs, read the trend and any point that breaks it. For bar charts, compare heights directly. The arithmetic is minimal; the skill is locating the one figure that bears on the pinned claim and ignoring the eye-catching figures placed elsewhere to distract you.

How do I define a claim precisely before evaluating choices?

Break the claim into four parts and state each one. The variable is the quantity or quality being asserted, such as a frequency, a growth rate, a temperature, a preference, or a count. The group is the subject the claim applies to, such as a species, a population, a sample, or a region. The condition is any qualifier the claim depends on, such as a time period, a treatment, or a comparison group. The direction is the shape of the assertion, such as higher or lower, increased or decreased. Say all four in one breath: “this species called more often after rain than during drought” becomes frequency, this species, after-rain versus drought, rising. Pinning this way forces you to notice the qualifiers that wrong answers exploit, like a comparison that excludes the middle period or a condition that rules out a later season. The discipline feels slow at first, but it is the single highest-value habit in the question type, because every elimination afterward flows from a precisely stated claim.

What is the right-topic-wrong-variable trap?

It is the most common distractor in command of evidence. The trap choice stays on the passage’s subject but quietly swaps the measured quantity for a related one. If the claim is about air temperature on tree-lined streets, the trap offers property values or pedestrian counts on those same streets, both genuinely about trees and streets, neither about temperature. Because the topic is intact, the choice feels relevant, and that feeling is exactly what the writers engineer. The defense is the variable check: name the claim’s exact quantity before you read the choices, and reject any choice that measures something else, however interesting or favorable it sounds. A practical habit is to underline the variable in your own restatement of the claim, so that when a choice presents a different noun you feel the mismatch at once. The deeper lesson is that the dangerous distractor is rarely the obviously irrelevant one; it is the one that keeps the topic and changes the measurement, since topic familiarity disguises itself as support.

How do I handle a “weakens the claim” variant?

Run the same two-step method, changing only the direction you require. Pin the claim’s variable, group, and condition exactly as for a support item, then look for the choice that engages those parts and contradicts the claim’s assertion. The correct weakening choice must do two things at once: match the claim’s subject and variable, and then push against it. A choice that points the right way but describes a different group or measures a different variable fails, because direction without a match is not opposition to this claim. Guard especially against the neutral choice, the true statement that takes no side on the claim’s variable; under pressure students treat any non-supporting choice as weakening, but a statement that neither props up nor knocks down the assertion sits outside the question. Say “weaken” to yourself before you screen so the requested direction stays active in working memory, and demand both engagement and contradiction from your selection. Done this way, the weaken variant is simply the support variant with the direction sign flipped.

How do textual and quantitative subtypes share the same logic?

They share it because the test is measuring one cognitive act in both: matching a specific claim to specific information in the requested direction. A sentence and a data point are interchangeable carriers of that information. Suppose a claim asserts a species called more often after rain. A textual item supports it with a sentence reporting more frequent calls after rain; a quantitative item supports it with a table row showing a higher call count in the rainy condition than the dry one. The reasoning that selects the right sentence is the reasoning that selects the right row: pin the claim, then choose the carrier that matches the variable, group, condition, and direction. Recognizing this unity consolidates your preparation. You stop learning two techniques and hoping the right one fires under pressure and start learning one technique that you apply to whichever surface appears. It also calms the common fear of the graphic, since the graphic is just a sentence written in numbers, screened the same way as any other choice.

How do I avoid a distractor from the wrong group or time?

Treat every group and every qualifier in the claim as load-bearing, because the writers specified them precisely so they could build a wrong answer just outside the boundary. When a claim names adults, expect a tempting choice about children; when it names this species, expect a choice about a related species; when it says after rainfall, expect a choice about a drought; when it says the first growing season, expect glowing data from a later season. After pinning the claim, scan each choice specifically for these boundary jumps before judging anything else about it. A choice that shifts the group fails even if its evidence is strong, and a choice that slips the time fails even if its variable and direction are right. A useful reflex is to flag the claim’s qualifiers as tripwires: if the claim says “after,” any “before” or “during” choice is suspect on sight and can often be eliminated before you finish reading it. The boundary is a wall, and evidence from the far side of the wall does not count.

Does a graphic always accompany quantitative evidence items?

Yes. The quantitative variant is defined by its pairing of a passage with a data display, and that display is what distinguishes it from the textual variant. The graphic is usually a table, a bar chart, or a line graph, and the passage sets up a claim that the data is meant to support, complete, or undercut. Without a graphic the item would be textual by definition. The presence of the display is also your cue to switch into label-first reading: before touching the numbers, read the title, the headers or axis labels, the units, and any note beneath the display, because those carry the qualifiers that decide close items. The graphic never demands real computation; comparing two values or reading whether a line rose or fell is the whole arithmetic. What it demands is accurate location of the one figure that bears on the pinned claim, while ignoring the prominently placed figures that answer neighboring questions. Read the structure first, pin the claim, then find the matching value last.

How do I compare two very similar evidence choices?

When two choices both survive a loose screen, the writers have built the distinction into one of the four claim parts, and you have not yet pinned the claim finely enough. Return to the claim and look for the exact variable, group, condition, or direction that separates the two survivors. Almost always the difference is in the variable: one choice restates the claim’s precise quantity while the other substitutes a broader or narrower cousin. A claim about short car trips, for instance, is matched by a choice about short car trips, not by a choice about overall congestion, which is a looser variable that could change for unrelated reasons. The resolution rule is that the correct choice fits the claim’s variable most exactly, and the loser broadens or narrows it slightly. Resist deciding by which choice sounds more impressive or more complete; tightness of match decides, not loudness. If you still cannot separate them, you are likely missing a qualifier in the claim, so reread the claim once more for a condition or comparison you skimmed.

What is the most common command of evidence mistake?

The most common mistake is choosing an answer that is on topic but addresses the wrong variable. A student reads the passage, recognizes the subject’s keywords in a choice, and selects it because the recognition feels like support, never checking whether the choice actually concerns the quantity the claim asserts. Closely behind it is reading the answer choices before pinning the claim, which lets the mind form a favorite and then rationalize toward it. Both errors share one root: impatience with the first step. Pinning the claim takes a deliberate beat that, under a ticking clock, feels unaffordable, so students skip it and pay for the skip in coin-flips between near-identical choices. The fix is to invest that beat every time, stating the claim’s variable, group, condition, and direction before reading any choice, and then to screen the choices mechanically against those four parts. Spend your time on the claim and almost none on liking or disliking choices, and the question type turns from fuzzy reading into reliable points.