Command of Evidence — Quantitative
Reading the graph for what it says, not for what the passage led you to expect.
~30 min · prequestion, worked examples, retrieval practice
A quantitative evidence question is two tests wearing one coat: is the statement true of the figure, and does the true statement do the job the sentence actually asks of it? Three of the four choices will pass one test and fail the other, which is why answering from the graph alone — or from the argument alone — loses the point with total reliability. It is also the most gettable hard question on Reading & Writing: everything needed is printed on the screen, with nothing to recall and nothing to interpret.
Before you read on
Two or three questions on exactly what this lesson teaches. Being wrong here is fine — it's the fastest way to find out what to pay attention to next.
Before any teaching. A text argues that new bus lanes increased ridership on the routes that received them. One choice reads: "Route 12, which did not receive a bus lane, carried more trips in 2023 than any other route." Assume that is an accurate reading of the graph. Can it be the credited answer?
A text reads: "In the greenhouse, the two fertiliser treatments produced nearly identical yields. In the field, however, ______" What must the data in the credited choice do?
A table reports that a district's literacy rate rose from 24 percent to 42 percent. Which statement describes that change accurately?
What this question actually asks
Every quantitative evidence item puts two things on the screen: a figure — a bar graph, a line graph, a scatterplot, or a table — and a short text, usually four to six sentences of science or social science, that says where the figure came from and makes a claim about it. The text either ends in a blank you complete, or is followed by a question asking which choice most effectively uses data from the figure to support, illustrate, or complete something specific.
The four choices are not four random facts. Typically two are accurate readings of the figure and two are not; on the hardest items all four are accurate. In every case exactly one does the job the text asks of it. That is the shape of the entire question type — accuracy and relevance are separate tests, and the item is engineered so that passing one does not get you past the other.
Where it sits on the test: College Board publishes Information and Ideas as roughly 26 percent of Reading & Writing, which is about twelve to fourteen of the fifty-four questions. It does not publish how those split between Central Ideas, Command of Evidence, and Inferences. Counts taken from released practice forms by the prep industry put the quantitative evidence items at roughly two to four per test — treat that as an estimate from published forms, not a College Board figure.
Foundations — reading a figure from zero (skip this block if it is already automatic)
Skip ahead if you can read an axis label without thinking about it. This block exists because the question type quietly assumes a set of conventions nobody is ever taught directly, and a student losing these points to the conventions rather than to the reasoning cannot tell the two apart from a score report.
Every figure has four pieces of furniture, and all four matter more than the bars do. The TITLE says what is being counted, and often for whom and when. The AXIS LABELS say what each direction measures and, critically, in what units. The SCALE says what one gridline is worth. The LEGEND says which bar, line, or dot belongs to which group. Read those four before you look at a single value.
Units are where points are lost silently. An axis reading "thousands of trips" means a bar at 40 stands for 40,000 trips. An axis reading "percent of households" means the numbers already are shares, so the difference between two of them is a difference in percentage points, not a percentage change. A column headed "cases per 100,000 residents" is a rate: the town with the most cases and the town with the highest rate are frequently different towns, and the question always wants whichever one the sentence named.
Table anatomy: the top row — the header — names what each column holds and gives its unit; every other row is one case, one town, one species, one year. To compare two cases, hold the column fixed and move down the rows. Most table errors are lateral, sliding one column across into numbers that look similar and mean something else.
Five phrases the test leans on that most students read straight past. "Increased BY 20" is a change of twenty; "increased TO 20" is a final value of twenty. "At least 40" includes forty; "more than 40" does not. "Twice as many" is a ratio; "20 more" is a difference. "Share" and "proportion" mean a rate; "number" and "total" mean a count. "Per capita" means divided by population, so it can rank places in the opposite order from the raw totals. None of this is vocabulary trivia — on this question type it decides which of two accurate-looking choices is the accurate one.
Last, the boundary that never moves: a figure can show what happened. It cannot show why. No graph, table, or scatterplot on the Digital SAT establishes a cause on its own, so a choice that explains rather than reports is wrong before you have checked a single number in it.
The method, in full
Step 1 — read the figure's furniture, not its values. Title, axis labels, units, scale, legend. Ten seconds. You are not extracting numbers yet; you are finding out what the numbers mean. This is the step that makes the unit traps impossible rather than merely avoidable.
Step 2 — read the text and find the claim the blank has to serve. Say it back in one clause: "the lanes brought riders back," "the small town punches above its size," "it was the light, not the soil." Then note the connective attached to the blank, because it constrains what the data must do. "However" and "by contrast" demand a difference from what was just said. "For example" and "for instance" demand one concrete case of the generalisation just made. "Because" and "since" demand the reason. "The data bear this out" demands direct confirmation of that claim — not of a neighbouring one.
Step 3 — predict, before reading any choice, what the data would have to show. Not the exact numbers: the shape. Which two groups get compared, which direction the difference runs, and roughly how large it must be to matter. This is the same discipline as predicting a word before reading the choices in Words in Context, and it works for the same reason — once you are holding a prediction, a plausible wrong choice has nothing left to be plausible against.
Step 4 — apply two filters, in this order. Filter one, IS IT TRUE OF THE FIGURE? Mechanical: find the row, the bar, the point; check the units; check the direction. A choice that fails here is dead and you never have to think about the argument at all. Filter two, DOES THE TRUE STATEMENT DO THE JOB? Does it compare the groups the claim is about, in the direction the claim needs, at the scope the claim specifies?
Truth first is deliberate, not arbitrary. Filter one is cheap and certain; filter two costs judgement. On a typical item filter one kills two choices in about fifteen seconds, so running the filters in the other order spends your judgement on statements that were never going to survive a units check. The weak path is to read four choices and pick the one that sounds like the claim. The strong path is to eliminate on arithmetic and decide on relevance.
Mechanism
Why the wrong answers are usually true
The item is built backwards from the figure, and that construction order determines what the distractors look like. A writer holding one graphic and needing three hard wrong answers has an almost free supply: every other comparison the same graphic supports is accurate, on topic, in the right register, and useless for the specific claim in the text. Manufacturing a false distractor takes deliberate work. Manufacturing a true-but-irrelevant one takes a second glance at a figure already drawn. That asymmetry is why "is it accurate?" is the weaker of the two filters even though it feels like the whole task — and it is why the choice you are drawn to is a liability. Your eye goes to the largest visual difference on the figure, and visual salience is a property of the drawing, not of the argument. The biggest bar is in the wrong answer for the same reason it is on the graph: because it is easy to see.
Worked examples
Fully worked — a table, and a claim about rates
- 01TABLE. "Rooftop solar installations, four towns, 2023." Columns: town; installations in 2023; total households. Alder: 480 installations, 12,000 households. Birch: 1,150 installations, 46,000 households. Cedar: 260 installations, 5,200 households. Dunmore: 900 installations, 30,000 households.
- 02TEXT. "Uptake of rooftop solar is not simply a function of town size. Cedar, the smallest town surveyed, ______"
- 03Furniture first: the table has two columns in different units — a count of installations and a count of households. Any claim about "uptake" is a claim about the ratio between them, not about either column on its own.
- 04Find the claim and the connective: the sentence sets uptake against town size and names Cedar. So the credited data must be about Cedar, and must show Cedar doing well on a measure that is not raw size.
- 05Predict, then check: Cedar should lead on installations per household. Cedar 260 ÷ 5,200 = 5.0%. Alder 480 ÷ 12,000 = 4.0%. Dunmore 900 ÷ 30,000 = 3.0%. Birch 1,150 ÷ 46,000 = 2.5%. Cedar is highest, and it is last on raw count.
- 06Run both filters on two candidates. "Cedar recorded the fewest installations of any town" — filter one: true, 260 is the smallest count. Filter two: it supports the opposite of the claim, so it dies. "Cedar installed solar on 5 percent of its households, a larger share than any larger town" — true, and it is exactly the size-independent measure the sentence set up.
- 07Answer: the rate statement, not the count statement. The count column is in the table only so that the wrong answer can be true.
One step hidden — a bar graph with a scaled axis
- 01FIGURE. Bar graph, "Annual research output at four institutes, 2010 and 2020." The vertical axis is labeled "hundreds of papers." Institute P: 12 in 2010, 15 in 2020. Institute Q: 30 in 2010, 24 in 2020. Institute R: 8 in 2010, 21 in 2020. Institute S: 26 in 2010, 27 in 2020.
- 02TEXT. "A national funding reform concentrated new grants at Institutes P and R after 2012, leaving Q and S on flat budgets. Output followed: ______"
- 03Furniture first: the axis is in hundreds, so Institute R's 2020 bar stands for 2,100 papers, not 21. Any choice that says "21 papers" has misstated the figure, whatever else it gets right.
- 04The claim: output rose where the funding went. The credited data must be about P and R, and is strongest if it also contrasts them with Q or S.
- 05Predict, then check: P and R both up, at least one of Q and S flat or down. P 12 → 15, up. R 8 → 21, up. Q 30 → 24, down. S 26 → 27, essentially flat.
Two steps hidden — two rival explanations
- 01TABLE. "Mean seedling height after eight weeks, in centimetres, under four conditions." Full light with fertiliser: 24.1. Full light, no fertiliser: 22.8. Shade with fertiliser: 9.6. Shade, no fertiliser: 8.9.
- 02TEXT. "Two explanations have been offered for the poor growth of seedlings under a closed canopy: too little light, or soil left nutrient-poor by the mature trees above. The results point to the first: ______"
- 03Furniture first: one measured quantity (height in centimetres) and two manipulated variables (light, fertiliser). Every comparison the table can support is a comparison of two of its four rows.
- 04The claim endorses one of two rival explanations, so the credited data must be a comparison the two explanations predict differently. Fertiliser moves height by under 1.5 cm at either light level (24.1 against 22.8; 9.6 against 8.9). Light moves it by roughly 14 cm at either fertiliser level (24.1 against 9.6; 22.8 against 8.9).
Solve alone
- 01FIGURE. "Share of surveyed commuters using each mode, City Z, 2015 and 2024," in percent of commuters surveyed. Car: 61 in 2015, 47 in 2024. Bus: 18, then 24. Bicycle: 4, then 9. Walking: 17, then 20. TEXT: "City Z's cycling investment is routinely credited with the fall in car commuting. The survey shows that cycling cannot account for most of it: ______" Work out what the credited statement has to compare, and what the arithmetic shows.
In your own words
In one sentence: why does Step 3 — deciding what the data would have to show before reading any choice — kill the accurate-but-irrelevant distractor, when reading the figure a second time does not?
Named traps
- True but irrelevant
- An accurate reading of the figure that supports a different claim from the one the sentence makes. This is the single most common wrong answer on the question type, and it is common by construction rather than by accident: every figure supports many comparisons, and only one of them is the one the text asked for.
- Superlative reflex
- Reaching for the largest or smallest value on the figure because it is the most visible thing on it, when the claim asked for a comparison between two named groups. A maximum is not a comparison. If the sentence does not contain a superlative, the answer almost certainly should not either.
- Count for rate
- Answering with a raw total when the claim is about prevalence, share, per capita, or per unit area — or with a percentage when the claim is about how many. The orchard with the most infected trees and the orchard with the highest infection rate are routinely different orchards, and both numbers are printed in the table so that both answers can be written.
- The printed number is not the quantity
- Quoting an axis value as if it were the amount. "Thousands of trips" makes a bar at 305 into 305,000. A rise from 24 percent to 42 percent is 18 percentage points and a 75 percent increase, and calling it "an 18 percent rise" misstates the data even though the subtraction was right. A choice that misstates the data cannot support anything, however well it fits the argument.
- Off-figure claim
- Asserting something the figure cannot contain: a cause, a motive, a reason, a value for a year that was not plotted, a group that is not in the legend, or a value more precise than the scale can carry. A figure reports what; it never reports why, and it never covers what it did not measure.
- Connective mismatch
- Data that continues where the sentence said "however," a generalisation where the sentence said "for example," or a contrast drawn on a different axis from the one the text raised — comparing two settings when the sentence was comparing two treatments. The connective is part of the question; it is not decoration on the sentence.
The 800-level margin
On the hardest items all four choices are accurate readings of the figure. Filter one eliminates nothing, and every second spent re-checking numbers is wasted. When that happens, stop reading the figure and start reading the claim's scope words — "only," "among first-year students," "after 2010," "relative to," "per capita," "more than doubled." The scope word is what makes exactly one of four true statements the answer, and it is almost always in the text, not the figure.
The rival-hypothesis variant is where strong students lose the point. The text offers two explanations and the claim endorses one. Data consistent with both explanations is not evidence for either, however large the effect and however accurately it is read — and the largest, most eye-catching comparison on the figure is usually the one both hypotheses predict, because it is the comparison between doing nothing and doing everything. The credited choice is the comparison the losing hypothesis cannot account for, which is often a smaller and visually duller difference.
Precision beyond what the figure supports is a defect, not a virtue. If a bar chart's axis is marked every 20 units and a bar sits between two marks, the figure supports "between 160 and 180" and does not support "exactly 175." Choices that quote an exact value off an unlabeled bar are wrong on the data even when they are right in spirit. A choice can also be too vague to serve — "Site 3 had many pairs" — but the Digital SAT builds the over-precise version far more often, because it looks like rigour.
A minority of items reverse the task: which finding would most directly weaken, undermine, or call into question the hypothesis. The two filters do not change — accurate, then relevant — but relevance now means the data must be something the hypothesis predicts should not happen. Students who have drilled the supporting form on autopilot answer the supporting question. The reversal is caught by reading the verb in the prompt, not by reading faster.
The execution errors that separate 1500 from 1600 here are all the same shape: the right operation on the wrong quantity. Right value, adjacent year. Right comparison, wrong group. A difference where the sentence asked for a ratio — "twice as many" is not "20 more than." A total where the claim was about a rate. A percentage-point change reported as a percent change. A sign flipped because the series was declining and the arithmetic was done as though it were rising. None of these is a reasoning failure; each is a failure to re-read the claim after doing the arithmetic, which costs four seconds and is the highest-yield habit available on this question type.
Time discipline, last. Read the figure's title, axis labels, units, and legend exactly once, before the text, and do not return to them. Nearly every wrong-row error happens on a second pass, when the eye goes back to the figure under time pressure holding a specific number and lands one line off.
Retrieval — with feedback on every choice
FIGURE. Bar graph, "Annual ridership by route, City Transit Authority, 2019 and 2023." The vertical axis is labeled "thousands of trips." Route 4: 820 in 2019, 610 in 2023. Route 9: 540 in 2019, 705 in 2023. Route 12: 1,150 in 2019, 980 in 2023. Route 17: 305 in 2019, 640 in 2023. TEXT. In 2020 the authority converted a full traffic lane into a dedicated bus lane on Routes 9 and 17; Routes 4 and 12 were left unchanged. Reviewing all four routes, an analyst argued that the ridership figures are consistent with the new lanes having produced a recovery: ______
Which choice most effectively uses data from the figure to complete the text?
COMMAND OF EVIDENCE - QUANTITATIVE - reference card Furniture first, once: title, axis labels, UNITS, scale, legend. Then never re-read them. Then the text: name the claim in one clause, and note the connective on the blank. Predict the shape of the needed data before reading a single choice. Filter 1 (cheap, mechanical): is it TRUE of the figure? Wrong units = wrong answer. Filter 2 (costly, decisive): does the true statement do the job the sentence asks? however / by contrast -> a difference. for example -> one concrete case. because / since -> the reason. "the data bear this out" -> that claim, not a neighbour. Count vs rate: most cases and highest prevalence are usually different rows. 24% to 42% = +18 percentage points = +75 percent. Two units, two numbers. "thousands of trips" on the axis: a bar at 305 is 305,000. A figure shows WHAT. Never WHY, and never a year it did not plot. Unlabeled bar between gridlines supports "between 160 and 180", not "exactly 175". Rival hypotheses: cite the comparison the losing one CANNOT explain, not the biggest gap. All four choices true? Stop reading the figure. Decide on the claim's scope words. Before locking: re-read the claim once, after the arithmetic. Right answer, wrong quantity is the 1600 leak.
Every item on this page is Meridian-original, written to match the Digital SAT's format and difficulty — it is not a real SAT question. The only source that matches the live test exactly is College Board's own Bluebook and Question Bank.
FIGURE. Bar graph, "Annual ridership by route, City Transit Authority, 2019 and 2023." The vertical axis is labeled "thousands of trips." Route 4: 820 in 2019, 610 in 2023. Route 9: 540 in 2019, 705 in 2023. Route 12: 1,150 in 2019, 980 in 2023. Route 17: 305 in 2019, 640 in 2023. TEXT. In 2020 the authority converted a full traffic lane into a dedicated bus lane on Routes 9 and 17; Routes 4 and 12 were left unchanged. Reviewing all four routes, an analyst argued that the ridership figures are consistent with the new lanes having produced a recovery: ______
Which choice most effectively uses data from the figure to complete the text?
- ARoute 12 carried more trips in 2023 than any other route.
Accurate — 980 thousand is the largest 2023 value on the figure — and it is the kind of statement the eye reaches for, because Route 12's bars are the tallest thing on the graph. But Route 12 never received a lane, and a single year's total says nothing about whether anything changed. It clears filter one and fails filter two.
- Routes 9 and 17, which received dedicated lanes, each carried more trips in 2023 than in 2019, while Routes 4 and 12, which did not, each carried fewer.
Correct. Check all four against the figure: Route 9, 540 → 705, up; Route 17, 305 → 640, up; Route 4, 820 → 610, down; Route 12, 1,150 → 980, down. ✓ It is the only choice that puts the treated routes and the untreated routes on the same axis of comparison, which is exactly what a claim about the lanes having produced a recovery requires.
- CRoute 4's ridership fell by more than 200,000 trips between 2019 and 2023, the largest change recorded on any route.
The first clause checks out — 820 to 610 is a fall of 210 thousand, and 210,000 is more than 200,000 — which is what makes the second clause easy to wave through. It is false: Route 17 changed by 335 thousand, a larger change than Route 4's 210. A choice whose opening clause verifies is not a verified choice.
- DRidership on Route 17 rose from 305 trips in 2019 to 640 trips in 2023.
The values are read off the correct bars, and the comparison is even relevant — Route 17 got a lane. The units are wrong. The axis is labeled "thousands of trips," so those bars stand for 305,000 and 640,000 trips. Being off by a factor of a thousand is misstating the data, and a misstatement supports nothing.
Traps tested: True but irrelevant · Half true composite · Unit blindness
TABLE. "Bacterial canker survey, four orchards, 2023." Columns: orchard; infected trees; trees surveyed. North: 96 infected of 4,800 surveyed. East: 150 of 3,000. South: 210 of 14,000. West: 44 of 1,100. TEXT. Treatment supplies were limited, so the regional agronomist recommended starting with the orchard where the disease was most prevalent — that is, where the largest share of surveyed trees was infected. The table identifies it: ______
Which choice most effectively uses data from the table to complete the text?
- ASouth, where 210 infected trees were recorded, more than at any other orchard.
The count is right and the measure is wrong. The sentence defines prevalence as a share, and South's 210 infections come out of 14,000 surveyed trees — 1.5 percent, the lowest rate in the table. The largest count sits in the orchard with the largest survey, which is why both columns are printed.
- BWest, where 44 of 1,100 surveyed trees were infected, a higher share than at North or South.
The method is right — this choice divides, as the sentence requires — and the comparison it makes is true: 44 ÷ 1,100 = 4.0 percent, against North's 2.0 and South's 1.5. It stops one orchard short. East is at 5.0 percent, so West is not the orchard the recommendation points to. A correct calculation on an incomplete scan.
- East, where 150 of 3,000 surveyed trees were infected — 5 percent, the largest share of any orchard.
Correct. Compute all four shares: North 96 ÷ 4,800 = 2.0%; East 150 ÷ 3,000 = 5.0%; South 210 ÷ 14,000 = 1.5%; West 44 ÷ 1,100 = 4.0%. ✓ East is highest, and it is second on raw count — precisely the gap between the two columns the item is built on.
- DNorth, where 4,800 trees were surveyed, more than at East or West.
True of the table, and about the wrong column entirely. Survey size is how hard the surveyors looked, not how sick the orchard is; North's 96 infections out of those 4,800 trees give it a 2.0 percent rate. Nothing about how many trees were examined bears on where the disease is most prevalent.
Traps tested: Count for rate · True but not superlative · True but irrelevant
TABLE. "Microplastic concentration by depth, sediment core from Lake Reyner," in particles per kilogram of dry sediment. Deeper layers were deposited earlier. 0–5 cm: 910. 5–10 cm: 350. 10–15 cm: 130. 15–20 cm: 40. TEXT. If the amount of plastic reaching the lakebed had been climbing at a steady rate, each successive layer would hold roughly the same amount more than the layer beneath it. Ecologist Hana Voss argues that the climb has instead been accelerating: ______
Which choice most effectively uses data from the table to support Voss's argument?
- AThe topmost layer holds 910 particles per kilogram, more than twenty times the 40 particles per kilogram in the deepest layer.
Accurate — 40 × 20 = 800, and the top layer holds 910 — and it is the most dramatic number pair available, which is why it is here. It establishes that deposition rose a great deal. It cannot distinguish a rising climb from a steady one, because a steady climb over four layers would also produce a large end-to-end difference. The text has already granted the increase; the argument is about its shape.
- Each step toward the surface is a larger increase than the step below it: 90 more particles per kilogram from the deepest layer to the next, then 220 more, then 560 more.
Correct. The text defines the steady-rate baseline as equal increases between layers, so the argument turns on whether those increases are equal. Compute them: 130 − 40 = 90; 350 − 130 = 220; 910 − 350 = 560. ✓ They grow at every step, which is exactly what an accelerating climb means and exactly what a steady climb rules out.
- CThe 15–20 cm layer holds 40 particles per kilogram, the lowest concentration measured anywhere in the core.
True, and it is a single value. An argument about the shape of a trend cannot be carried by one measurement, because one measurement contains no rate of change — it needs at least two differences set against each other. This choice would be equally at home under the steady-rate hypothesis the text is arguing against.
- DConcentration declines steadily with depth, falling by roughly the same amount from each layer to the one below it.
It declines with depth, so the first half sounds right, and the second half is false and fatal. The falls are 560, then 220, then 90 — nowhere near equal. Worse, "roughly the same amount" is the text's own description of the steady-rate hypothesis, so even if it were true it would support the position Voss is arguing against.
Traps tested: Supports weaker claim · Single point no comparison · Contradicts the claim
TABLE. "Germination of Ceanothus seeds after simulated fire treatments," as the percentage of seeds germinating. Untreated control: 4. Heat only: 31. Smoke only: 9. Heat and smoke together: 34. TEXT. Chaparral shrubs germinate en masse in the season after a wildfire, and two cues have been proposed for it: the pulse of heat the fire drives into the soil, and the chemical compounds carried in its smoke. Botanist Ines Arrieta concludes that for Ceanothus the cue is heat: ______
Which choice most effectively uses data from the table to support Arrieta's conclusion?
- ASeeds exposed to both heat and smoke germinated at 34 percent, compared with 4 percent for untreated seeds.
Accurate, on topic, and the largest contrast in the table — and it is the one comparison both hypotheses predict. If smoke were the cue, seeds given heat and smoke would also far outperform untreated seeds. Data that both rival explanations predict is not evidence for either of them, whatever its size.
- BSeeds exposed to smoke alone germinated at 9 percent, more than twice the 4 percent recorded for untreated seeds.
The arithmetic holds — twice 4 is 8, and smoke-only reached 9 — but it points the wrong way. This is the effect of smoke on its own, which is the evidence the rival hypothesis would cite. It counts against Arrieta's conclusion rather than for it, and the smallness of the effect is a separate matter from its direction.
- Seeds exposed to heat alone germinated at 31 percent, against 9 percent for seeds exposed to smoke alone.
Correct. This is the only comparison in the table that isolates one cue against the other with everything else held constant, and the two hypotheses predict opposite results for it. Heat alone reaches 31 percent — close to the 34 percent from both cues together — while smoke alone reaches 9 percent, close to the 4 percent baseline. ✓ Nearly all of the effect travels with the heat.
- DUntreated seeds germinated at 4 percent, the lowest rate of any group in the study.
True, and it is the control, so it feels foundational. It establishes only that the treatments did something — the claim at issue is which treatment. A statement about the untreated group cannot separate two treatments from each other, because it involves neither of them.
Traps tested: Consistent with both hypotheses · Supports rival hypothesis · True but irrelevant
TABLE. "Households with home internet access, four districts," in percent of households. District A: 41 in 2016, 62 in 2023. District B: 68 in 2016, 79 in 2023. District C: 22 in 2016, 51 in 2023. District D: 84 in 2016, 88 in 2023. TEXT. Reporting on the period, a regional authority stated that the distance between its least-connected district and its best-connected district had closed considerably. The figures bear this out: ______
Which choice most effectively uses data from the table to complete the text?
- ADistrict C's access rate rose by 29 percent, the largest increase recorded in any district.
Both halves are almost right, and "almost" is fatal. District C did post the largest rise, and that rise is 29 percentage points, not 29 percent: as a percent increase it is 29 ÷ 22, about 132 percent. A difference between two percentages is measured in percentage points, and a choice that misstates the unit has misstated the data.
- In 2016 District D exceeded District C by 62 percentage points; by 2023 that difference had fallen to 37 percentage points.
Correct. District C is the least-connected district in both years and District D the best-connected in both, so those are the two the sentence is about. 2016: 84 − 22 = 62 percentage points. 2023: 88 − 51 = 37 percentage points. ✓ The distance closed by 25 points, which is what "closed considerably" needs and what no other choice reports.
- CDistrict D had the highest access rate in both 2016 and 2023.
True on both counts — 84 leads in 2016 and 88 leads in 2023 — and it identifies one of the two districts the claim is about. It reports that the ranking held, and the claim is about the size of a gap. A gap can narrow enormously while the order at the top never changes, so this statement is compatible with the claim and does not support it.
- DBy 2023, District B had a higher access rate than District A.
Accurate: 79 against 62. Wrong pair. The sentence is about the distance between the least-connected and the best-connected districts, which are C and D; A and B are the two middle districts and their ordering says nothing about the extremes.
Traps tested: Percent vs percentage points · True but irrelevant · Wrong groups compared
FIGURE. Bar graph, "Nesting pairs of piping plover at four monitored beaches, 2014 and 2024." The vertical axis is marked at intervals of 20 pairs, and no bar carries a printed value. Site 1: the 2014 bar rises just past the 120 mark, the 2024 bar stops just short of 60. Site 2: 2014 just past 40, 2024 just past 100. Site 3: the two bars stand level with each other, a little below 180. Site 4: 2014 just short of 100, 2024 just past 20. TEXT. Sites 1 and 4 were opened to off-road vehicle traffic in 2016; Sites 2 and 3 remained closed to it throughout. A monitoring biologist notes that the counts are consistent with — though on their own they cannot demonstrate — a harmful effect of vehicle access: ______
Which choice most effectively uses data from the figure to complete the text?
- AVehicle traffic reduced nesting success at Sites 1 and 4.
This states the conclusion instead of the evidence for it, and the text has just ruled it out in its own words: the counts cannot demonstrate a harmful effect on their own. Two sites falling while two others did not is consistent with a cause and does not establish one — a figure reports what happened, never why.
- BNesting pairs at Site 1 fell below 60 for the first time in 2016, the year vehicle access began.
It sounds like the strongest possible evidence, which is the point: it names the mechanism year. The figure plots two years, 2014 and 2024, and nothing in between. There is no bar for 2016, so the claim is unreadable from this figure whether or not it is true.
- CSite 3 recorded exactly 175 nesting pairs in both 2014 and 2024.
The unchanged pair of bars at Site 3 is real and relevant, and the number attached to it is not. The axis is marked every 20 pairs and the bars are unlabeled, so a bar sitting a little below 180 supports "between 160 and 180" and cannot support "exactly 175." Precision beyond what the scale carries is a misreading, not a sharper reading.
- Both sites opened to vehicles ended the period with fewer than half the pairs they began with, while neither closed site declined at all.
Correct, and every part of it survives a bounded reading of the scale. Site 1 began above 120, so half is above 60, and it ended below 60. Site 4 began between 80 and 100, so half is between 40 and 50, and it ended between 20 and 40. ✓ Site 2 rose, from just past 40 to just past 100, and Site 3 held level. It uses the figure at exactly the precision the figure supports and puts the opened sites against the closed ones, which is the contrast the claim rests on.
Traps tested: Causal upgrade · Unplotted value · Over precision beyond scale
Up next
Inferences
The most-missed type on the section. The bar is logical necessity, not plausibility.
28 min