What Can the Evidence Really Prove?

(Original paired practice passage and dataset. Place names, pilot details, and numerical results are constructed for test-preparation use.)

Cities facing rising summer temperatures have no shortage of proposed fixes, but few deserve the priority currently being given to expanding urban tree canopy. Reflective roofing, permeable pavement, and misting stations all have their advocates, and each can play a supporting role. None of them, however, should be allowed to crowd out what the evidence increasingly suggests is the single most cost-effective long-term intervention available to city planners: getting more trees into the ground, and keeping them there.

The strongest evidence for this position comes from Fairhaven's four-year pilot program, which tracked peak summer surface temperatures across fifty-six neighborhoods, fourteen in each of four intervention groups. Neighborhoods that received expanded tree canopy alone saw meaningfully larger average temperature reductions than those that received reflective roofing alone. Trees, unlike reflective surfaces, do more than bounce heat away; they actively cool the surrounding air through evapotranspiration, provide shade that reduces the temperature of the pavement and buildings beneath them, and continue delivering these benefits for decades after the initial planting cost has been paid.

Critics of canopy-first strategies typically raise two objections. The first is cost: a mature tree canopy takes years to establish, and newly planted trees require irrigation, protection from pests, and replacement when they fail to survive, all of which carry ongoing expense that reflective roofing, installed once, mostly avoids. The second objection is reliability: unlike a roof coating, a tree is a living thing, vulnerable to drought, disease, and the simple bad luck of an unusually harsh season, which makes canopy programs a less predictable investment than manufactured alternatives.

Both objections have real force, and this essay does not dismiss them. Canopy programs are genuinely more expensive to establish and maintain in their early years than reflective roofing, and their outcomes are genuinely less uniform, as even Fairhaven's own pilot data show. But the appropriate response to a slower, less predictable investment with a larger long-term payoff is not to abandon it in favor of a faster, more uniform one with a smaller payoff; it is to budget for the years the investment takes to mature, and to treat canopy expansion as what it actually is: an infrastructure commitment, not a one-time purchase. Cities that have made this shift in framing, treating trees as the long-term backbone of a heat-reduction strategy rather than a decorative supplement to more "serious" engineering solutions, are the cities most likely to still be seeing returns on that investment decades from now.

I have spent the better part of a decade designing reflective and green roofing systems for municipal buildings, and I want to say plainly, before anything else: I am not opposed to trees. I am opposed to a framing, increasingly common in city planning circles, that treats tree canopy as the default answer to urban heat and every other intervention as a secondary supplement to it. That framing gets the engineering backward.

Reflective roofing has one significant advantage that no amount of canopy enthusiasm should obscure: it works immediately, uniformly, and predictably. A newly coated roof begins reflecting solar radiation the day the coating is applied, at a rate that varies little from building to building and does not depend on soil quality, water availability, or the survival of a young tree through its vulnerable first several years. For a city trying to reduce heat-related illness this summer, not in fifteen years, that immediacy is not a minor convenience; it may be the difference that matters most.

None of this means canopy expansion is a poor investment; the same Fairhaven data that canopy advocates cite shows real, if more variable, cooling benefits from tree planting, and I have no quarrel with the finding that combining both interventions produced the largest reductions recorded in the pilot. My disagreement is narrower than a wholesale rejection of trees: it is with the specific claim that canopy expansion deserves to be treated as the primary strategy, with other interventions relegated to supporting roles. The Fairhaven data, read carefully, supports a different conclusion, one its canopy-focused advocates tend to underemphasize: the combined approach outperformed either intervention alone by a wide margin, which suggests that trees and reflective surfaces are functioning as complements, not as a hierarchy with one clearly superior member.

A city with the patience and budget to wait fifteen years for a mature canopy, and the resources to maintain reflective roofing in the meantime, should absolutely do both. But a city forced to choose where its next dollar goes, this year, facing a heat emergency it cannot wait out, should not be told that the "real" solution is the slower one. Immediacy is not a lesser virtue than durability; it is a different one, suited to a different part of the problem, and a genuinely evidence-based heat strategy would treat both interventions as answering different questions rather than competing for the same crown.

Figure 1: Compare the visual with the claims in both passages.

Identify the claim, who owns it, and what pushes against it.

Before evaluating whether a passage's evidence is any good, identify exactly what claim that evidence is supposed to support. Evidence cannot be judged in a vacuum; the same piece of evidence can be strong support for one claim and weak or irrelevant support for a different, nearby claim, which means naming the claim first is not a preliminary formality but the step that makes evidence evaluation possible at all.

Separate the subject from what the author is claiming about it. A topic names what a passage is about; a claim takes a position or makes an assertion about that topic, something that could, in principle, be argued against. "Urban tree canopy" is a topic. "Urban tree canopy deserves priority over other heat-reduction strategies" is a claim. Confusing the two produces answers that correctly identify what a passage discusses without identifying what it actually argues, which is a different and more demanding task.

Find the position the writer is directly trying to establish.

Passage E states directly: "few deserve the priority currently being given to expanding urban tree canopy... None of them, however, should be allowed to crowd out what the evidence increasingly suggests is the single most cost-effective long-term intervention available to city planners."

The topic is urban heat-reduction strategies. The central claim, stated directly, is more specific and arguable: tree canopy expansion deserves priority over competing interventions as the primary long-term strategy. A choice describing the passage's topic as "urban heat reduction methods" would be accurate but would not answer a question asking for the passage's central claim.

Distinguish the central claim from a supporting claim. A passage's central claim is the position the whole argument is built to support; a secondary or supporting claim is a smaller assertion that helps build the case for the central claim but is not, itself, the passage's main point. A secondary claim can sometimes sound broader or more dramatic than the central claim, which is exactly what makes mistaking one for the other a persistent trap.

Separate a supporting claim from the passage's main claim.

Passage E asserts, in support of its central claim, that trees "actively cool the surrounding air through evapotranspiration" and "continue delivering these benefits for decades after the initial planting cost has been paid."

This is a secondary claim: a specific assertion about why trees are effective, offered in service of the passage's central claim that canopy expansion deserves priority. A reader who mistakes this vivid, specific claim about evapotranspiration for the passage's central claim has swapped a supporting point for the main argument it was built to support.

Track who owns each claim. Every claim in a passage belongs to someone: the author, a quoted expert, an opposing group, a cited study, or, in a paired-passage set, one specific author rather than the other. Correctly attributing a claim to its actual owner, rather than assuming every claim belongs to the passage's author, is required before that claim can be evaluated or compared to anything else.

Keep each claim attached to the source that actually makes it.

Passage E states that critics raise concerns about cost and reliability, specifically that "a tree is a living thing, vulnerable to drought, disease, and the simple bad luck of an unusually harsh season."

This claim belongs to the canopy program's critics, not to Passage E's author, who reports the objection accurately and even grants that it "has real force" before ultimately arguing that it does not outweigh the long-term case for canopy investment. A question asking what Passage E's author believes about tree reliability should not be answered using this sentence alone, since it represents an objection the author is reporting and responding to, not asserting as their own view.

Find the counterclaim before following the response to it. A counterclaim is an opposing position the author acknowledges, and it has to be identified clearly before evaluating how the author responds to it: whether the author rebuts it, concedes part of it, qualifies it, or leaves it unresolved. Skipping past the counterclaim itself and jumping straight to the author's response risks misunderstanding what, exactly, is being conceded or rejected.

Separate a counterclaim from the author's own position. A counterclaim belongs to whoever raises it, typically an opposing voice the author is representing fairly before responding, and it should never be mistaken for a position the author personally holds, even when the author describes that counterclaim in detailed, sympathetic terms. An author can present an opposing view accurately and even generously without endorsing it.

Follow how the argument changes after the counterclaim. An argument does not always simply reject a counterclaim outright; it can concede part of the counterclaim's force while maintaining its overall position, and tracking exactly what gets conceded and what remains defended is necessary to understand the argument's actual final shape, rather than assuming every counterclaim ends in total refutation.

Identify the objection first, then ask how the writer answers it.

Passage E explicitly states, "Both objections have real force, and this essay does not dismiss them. Canopy programs are genuinely more expensive to establish and maintain... and their outcomes are genuinely less uniform," before concluding, "the appropriate response... is not to abandon it... it is to budget for the years the investment takes to mature."

A reader who assumes the author must fully refute the counterclaim, rejecting the claims about cost and unpredictability as false, would misread this passage; the author explicitly concedes both points are genuinely true. What the author rejects is not the truth of the objections but the conclusion that those objections should be drawn from them, that the cost and unpredictability should lead a city to abandon canopy investment rather than budget appropriately for it. The argument's real shape is partial concession followed by a redirected conclusion, not outright denial.

Ask what each piece of evidence actually proves.

Before deciding whether a piece of evidence supports a claim, state plainly what that evidence, on its own, actually establishes. This is a check against the common failure of noticing that evidence sits near a claim and assuming, without further checking, that it therefore supports that specific claim.

Ask which claim this evidence is meant to support.

Passage E cites the finding that "neighborhoods that received expanded tree canopy alone saw meaningfully larger average temperature reductions than those that received reflective roofing alone."

On its own, this evidence establishes a comparison between two single interventions, canopy alone versus roofing alone, in this specific pilot program. It does not, by itself, establish that canopy expansion is more cost-effective, more reliable, or superior to a combined approach; those are separate claims that would each require their own supporting evidence, not automatically supplied by this one comparison.

Keep relevant evidence tied to the exact claim. Evidence is relevant only when it actually bears on the specific claim being evaluated, not merely when it is true, interesting, and generally related to the passage's topic. A piece of evidence can be entirely accurate and still fail to support the exact claim a question is asking about.

Related evidence is not useful unless it bears on the exact claim.

If a question asked what evidence Passage F offers for the claim that reflective roofing "works immediately," the detail that Passage F's author has "spent the better part of a decade designing reflective and green roofing systems" would be true and related to the passage's general subject, but it is a credential establishing the author's expertise, not evidence establishing the immediacy of the roofing's effect. The actual evidence for immediacy is the specific claim that follows: a coated roof "begins reflecting solar radiation the day the coating is applied."

Separate strong evidence from a large amount of evidence. The quantity of evidence a passage offers is not, by itself, a measure of that evidence's strength, and a claim supported by several weak or loosely related pieces of evidence is not thereby better supported than a claim backed by one precise, directly relevant piece. Evaluating evidence requires checking its quality and its fit to the claim, not counting how many sentences are devoted to it.

Ask whether the evidence is enough for the size of the claim. Evidence has to be evaluated not only for whether it is relevant and sound but for whether its scope and amount actually justify the scope of the claim it is being used to support. Evidence that is entirely relevant and entirely accurate can still be insufficient if the claim it supports is broader than what that evidence, by itself, can establish.

Ask whether the evidence is enough for the size of the claim.

Passage F cites the Fairhaven pilot's finding about combined interventions to support the claim that "trees and reflective surfaces are functioning as complements, not as a hierarchy with one clearly superior member."

The pilot data, drawn from one city's fourteen neighborhoods over four years, is relevant and directly on point for this specific comparative claim. Whether that same evidence would be sufficient for a much broader claim, that trees and reflective surfaces function as complements in cities generally, is a separate question; a single pilot program's data may be sufficient for a claim scoped to that pilot's own findings while still being insufficient, on its own, for a sweeping generalization beyond it.

Notice when the evidence supports only part of the claim. A multi-part claim can be only partially supported by the evidence a passage offers, with one component well established and another component simply asserted alongside it without direct support. Recognizing this partial coverage prevents treating an argument as fully proven when only a portion of it actually is.

Use the passage to test the argument's reasoning. Evaluating an argument's reasoning means checking whether its conclusion actually follows from its stated evidence and premises, not whether the conclusion sounds reasonable or matches a reader's own prior views. A reasoning flaw is a gap between what the evidence establishes and what the argument claims that evidence proves, and finding that gap requires comparing the two directly.

Find the unsupported step between the evidence and the conclusion.

Suppose a version of Passage E argued that because canopy-only neighborhoods outperformed roofing-only neighborhods, cities should abandon reflective roofing programs entirely.

The evidence, a comparison between canopy-alone and roofing-alone outcomes, partly supports a claim about relative effectiveness between those two specific options. It does not support the broader conclusion that roofing programs should be abandoned entirely, since the same data, and Passage E's own actual argument, shows combined interventions outperforming either single approach, undermining rather than supporting an "abandon roofing" conclusion. The premises here would partly support the conclusion's general direction while leaving an unsupported logical bridge to its most extreme version.

Strengthen the exact claim the question names. A strengthen question requires identifying precisely which claim the question is asking about and what new information would make that specific claim more likely to be true, based on the inferential link the argument already depends on. New information that supports a different, related claim, without touching the actual claim named in the question, does not strengthen the argument the question is asking about.

Choose new information that makes the exact claim more likely.

Suppose new information revealed that Fairhaven's combined-intervention neighborhoods also reported the lowest rates of heat-related hospital visits during the pilot period.

This new information would most directly strengthen Passage F's specific claim that combined interventions function as complements producing the strongest outcomes, since it adds an additional, independent measure, health outcomes rather than just surface temperature, pointing toward the same conclusion. It would do comparatively little to strengthen Passage E's narrower claim that canopy alone deserves priority over other interventions, since the new information specifically involves the combined approach, not canopy alone.

Weaken the link the argument depends on. A weaken question is most effectively answered by identifying the specific inferential link an argument depends on, the exact step connecting its evidence to its conclusion, and finding new information that damages that link specifically, rather than information that must first be interpreted as proving the argument's conclusion is the opposite of true.

Choose new information that damages the link between evidence and claim.

Suppose new information revealed that Fairhaven's reflective roofing was replaced with a new, more effective coating partway through the pilot, meaning the reflective-roofing group's results reflect roughly half the years using an older, less effective product.

This would weaken Passage E's implicit comparison between canopy and roofing effectiveness, not by proving canopy is ineffective, but by damaging the reliability of the specific comparison the author's evidence depends on: if the roofing group's own results were artificially lowered by using an outdated coating for part of the study, the reported gap between canopy and roofing outcomes may not reflect a fair comparison between the two interventions as they currently exist. The new information does not need to prove reflective roofing is actually superior; it only needs to undermine the specific comparison the argument relies on.

Distinguish a fact from an opinion. A fact is a statement presented as verifiable information, something that could, in principle, be checked against external evidence and confirmed or disconfirmed. An opinion is a belief or evaluation that is not itself established as fact, even when it may be widely shared or strongly felt. The presence of a number in a statement does not automatically make it a fact, and the absence of one does not automatically make a statement an opinion.

Separate a verifiable statement from a personal judgment.

Passage E states that Fairhaven's pilot program "tracked peak summer surface temperatures across fourteen neighborhoods," a verifiable, factual claim about what the program did, distinct from the passage's evaluative claim that tree canopy "deserves priority" over other interventions, which is a position the passage argues for rather than a fact it reports.

A reader who treats every numerically specific statement in Passage E as equally factual, without distinguishing the verifiable details of the pilot's design from the passage's own evaluative argument built on top of those details, will have trouble separating what the passage establishes from what it is trying to persuade the reader of.

Recognize a reasoned judgment that uses evidence. A reasoned judgment is a distinct category from both fact and pure opinion: it is an evaluative conclusion that is grounded in evidence and reasoning, rather than either a bare verifiable fact or an unsupported preference. Treating every evaluative-sounding statement as a mere opinion, with no distinction for whether it is actually backed by evidence, collapses a distinction this test specifically tests.

A judgment can be evidence-based without becoming a fact.

Passage F's claim that "the combined approach outperformed either intervention alone by a wide margin, which suggests that trees and reflective surfaces are functioning as complements, not as a hierarchy" is not a bare, verifiable fact, since it involves interpretation, and it is not a pure opinion untethered from evidence, since it is explicitly grounded in the pilot's comparative data.

This is a reasoned judgment: an evaluative conclusion, but one built directly on cited evidence rather than asserted as a simple preference. A fact/opinion question that offered only "fact" or "opinion" as choices for this statement would be testing exactly the oversimplified binary this mechanism is built to move past.

Compare the sources only after each one is clear.

Before comparing two passages, understand the first one entirely on its own terms: its central claim, its supporting evidence, its counterclaims, and its author's specific position. This applies to whichever passage a paired-passage question set presents first, regardless of how it happens to be labeled. Treating the two passages as a blended, undifferentiated whole from the very first read is what causes claims from one passage to get attributed to the other.

Read Passage B on its own first. The same discipline applies to the second passage: map its central claim, evidence, and position independently, without yet asking how it relates to the first. Only once both passages are individually clear does comparing them become a reliable process rather than a guess based on a vague, blended impression of "what both passages were generally about."

Keep each passage's claims and evidence separate. Throughout this process, keep track of which passage each specific claim, piece of evidence, or counterclaim actually belongs to. A detail from Passage E should not casually migrate into your understanding of what Passage F argues, even when the two passages address closely related material, and confusing the two sources is one of the most common errors on cross-text questions.

Name the relationship before comparing the passages. Once both passages are individually clear, state in your own words what kind of relationship connects them before looking at any answer choices: full agreement, outright disagreement, partial agreement with qualification, an extension of one point by the other, a shared conclusion reached through different evidence, or a difference in emphasis rather than substance. Naming this relationship first prevents defaulting to a simple agree-or-disagree framing that may not actually fit.

Find where the passages agree. Identifying genuine points of agreement between two passages requires locating a specific claim both authors actually affirm, not simply noticing that both passages address the same general subject.

State the exact point on which the two passages agree or differ.

Both Passage E and Passage F affirm that Fairhaven's combined intervention, tree canopy plus reflective roofing, produced the largest temperature reductions recorded in the pilot program.

This is genuine agreement: both authors cite and accept this specific finding, even though they go on to draw different emphases and recommendations from it. A comparison question asking where the two passages agree should point to this shared acceptance of the combined-intervention finding, not to the broader topic of urban heat, which both passages address but do not necessarily agree about in every respect.

Find the exact point where the passages differ. Locating a genuine disagreement between two passages requires identifying the specific claim on which the authors actually take different positions, distinguished from surrounding material both authors might agree on.

Compare how the two authors or narrators view the same issue.

Passage E's author argues that tree canopy expansion "deserves priority" as the primary long-term heat-reduction strategy, while Passage F's author explicitly objects to "a framing... that treats tree canopy as the default answer... and every other intervention as a secondary supplement to it."

This is the exact point of difference: not whether canopy expansion is valuable, which both authors accept, but whether it deserves to be treated as the primary, priority strategy rather than one of two complementary approaches. Locating this precise point of difference is more useful than a vague sense that "the authors disagree about trees," which does not capture what they actually disagree about.

Recognize partial agreement and qualification. Two passages often relate to each other through partial agreement or qualification rather than a clean binary of agreeing or disagreeing, with one author accepting most of another's claim while narrowing, complicating, or adding a condition to it. Treating every cross-text relationship as either full agreement or full disagreement misses this common and specifically tested middle case.

Notice when a source partly agrees while limiting the claim.

Passage F's author states directly, "None of this means canopy expansion is a poor investment... My disagreement is narrower than a wholesale rejection of trees."

This is qualification, not disagreement: Passage F's author accepts the value of canopy expansion and even the specific Fairhaven findings Passage E cites, while narrowing the disagreement specifically to whether canopy deserves top priority over other interventions rather than equal, complementary status. A choice describing the two passages as being in fundamental disagreement about the value of tree canopy would miss this qualification and overstate how far apart the two authors actually are.

Use one passage to test a claim made in the other. One passage's specific evidence or argument can be used to evaluate a claim made in the other passage, testing whether the first passage's material actually supports, challenges, or is simply neutral toward the second passage's position. This relationship can be indirect, requiring a reader to connect a detail in one text to a claim in the other rather than finding an explicit cross-reference.

Use one passage only when it actually provides evidence for the other passage's claim.

Passage F's specific detail that reflective roofing "begins reflecting solar radiation the day the coating is applied" indirectly challenges an implicit assumption in Passage E's argument, that a slower, longer-term investment is worth prioritizing over faster alternatives, by supplying a concrete reason immediacy might matter for reasons Passage E's argument does not directly address, such as an ongoing heat emergency a city cannot wait out.

Passage E never mentions this specific point directly, and Passage F never explicitly states that it is responding to Passage E's argument about long-term priority; the connection has to be made by the reader, using Passage F's specific evidence to test the strength of an assumption underlying Passage E's case.

Combine evidence from both passages only after each source is clear. A full cross-text synthesis question requires drawing on specific evidence from both passages together, not answering from just one source while ignoring the other, and this kind of synthesis should only be attempted once each passage's individual claims and evidence are already clearly mapped.

Draw only the conclusion that becomes supportable when both sources are combined.

A question asking what conclusion is best supported by considering both passages together might be answered: the Fairhaven pilot data, as reported consistently across both passages, supports treating tree canopy and reflective roofing as complementary interventions whose combination outperforms either alone, even though the two authors disagree about which single intervention, if a city can only afford one, should receive priority.

This synthesis draws specifically on Passage E's data about the combined intervention's results, Passage F's explicit framing of the two interventions as complements, and both authors' shared acceptance of the combined finding, while still accurately representing the genuine point of disagreement that remains between them. An answer drawing only on Passage E, or only on Passage F, would not be a true synthesis, however accurate it was as a description of one passage alone.

Read the visual at the same scope and precision as the text.

Before interpreting what a graph shows, read its title and its axis labels first, since these establish exactly what is being measured, over what period, and for what population, information that controls every subsequent interpretation of the visual's shape.

Check the axes, units, categories, and legend. A visual's axes specify the scale and units the data is measured in, its categories specify what is being compared, and its legend, when present, specifies what each color, pattern, or symbol represents. Skipping this step and jumping straight to the visual's overall shape risks misreading what the visual actually displays, especially when an axis does not start at zero or uses an unfamiliar scale.

Describe the visual before connecting it to the passage. Before asking what a visual means for a passage's argument, describe plainly what the visual itself shows: which category has the highest or lowest value, what the general pattern is, and whether any data point breaks from that pattern. This description should be possible without reference to the passage at all, since the visual has its own content independent of any text describing it.

Read the figure before using the passage to interpret it.

Figure 1 shows that the combined intervention produced the largest average reduction in peak surface temperature, 9.5°F, compared to 6.2°F for tree canopy alone, 4.5°F for reflective roofing alone, and 0°F for no intervention.

This description uses only the figure itself: its labeled values and its categories, without yet drawing any connection to what either passage argues about those results. A visual-alone question can be answered directly from this kind of description, without needing to consult the surrounding text at all.

Match the graph to the exact claim in the text. Once the visual is described on its own terms, the next step is identifying precisely which claim in the passage the visual is being used to evaluate, since a visual can be relevant to one specific claim in a passage while being only loosely connected, or entirely unconnected, to other claims the same passage makes.

Notice when the visual confirms the passage. A visual confirms a textual claim when the data it displays directly matches what the passage asserts, using the same variable, population, and scope the passage is actually claiming something about.

Check whether the figure measures the exact claim made in the text.

Passage E's claim that canopy-only neighborhoods "saw meaningfully larger average temperature reductions than those that received reflective roofing alone" is directly confirmed by Figure 1's reported values: 6.2°F for canopy alone compared to 4.5°F for roofing alone.

The figure's specific values, for the specific two categories the passage's claim compares, match what the passage asserts, making this a clear case of visual confirmation rather than a case requiring careful qualification.

Notice when the visual qualifies the passage. A visual can support a passage's general claim while also narrowing, complicating, or adding a condition to it, revealing detail the prose alone does not convey. A visual that qualifies a claim is not contradicting it; it is adding precision or an exception the text's general statement leaves out.

Decide whether the visual weakens, limits, or directly contradicts the claim.

Passage E describes canopy programs' outcomes as "genuinely less uniform" than reflective roofing's, and Figure 1 qualifies this general claim with a specific detail: the wider error bars for the tree-canopy-only group than for the roofing-only group, together with the separately noted Riverside result of 2.1°F, add specific evidence that canopy outcomes varied more across neighborhoods in this constructed pilot.

This is qualification, not contradiction: the figure supports the passage's general claim about variability while adding a specific, textually unstated detail, exactly which neighborhood underperformed and roughly why, that sharpens rather than undermines the passage's point.

Notice when the visual contradicts the passage. A visual contradicts a passage when the data it displays does not match what the passage claims about that same variable, population, and scope, and recognizing a genuine contradiction requires checking that the visual and the text are actually measuring the same thing before concluding they disagree.

Check whether the text and visual cover the same population, time, and variable.

If a passage claimed that reflective roofing "outperformed all other single interventions" in the Fairhaven pilot, Figure 1 would directly contradict that specific claim, since it shows tree canopy alone, at 6.2°F, outperforming reflective roofing alone, at 4.5°F, on the same measure, over the same period, for the same pilot program. Before concluding this is a genuine contradiction, however, a careful reader would confirm that the passage's claim and the figure's data are describing the same variable, peak surface temperature reduction, and the same scope, single interventions only, rather than assuming a contradiction where the text and visual are actually measuring different things.

Keep correlation separate from causation. A visual showing that one variable rises alongside another establishes an association, not a proven causal mechanism, unless the passage's text or the visual's own design specifically establishes causation, for instance through a controlled comparison. Treating a visual trend as automatic proof of a causal claim beyond what the data or text actually support is a specific and commonly tested error.

Do not turn an association in the data into proof of cause.

Figure 1 shows an association between intervention type and temperature reduction in the Fairhaven pilot. The passage does not state that neighborhoods were randomly assigned or otherwise controlled in a way that would isolate intervention type as the cause. The figure therefore supports a comparative association within this pilot, not a definitive causal claim.

A broader causal claim, such as asserting that tree canopy expansion would produce an identical, guaranteed temperature reduction in any city regardless of climate, soil, or baseline conditions, would overreach beyond what even this well-designed pilot's data can support on its own, since the study's causal design is scoped specifically to Fairhaven's own conditions during this particular four-year period.

Match the precision of the answer to the precision of the data. An answer choice should claim no more precision than the visual actually supports, and when a visual displays approximate values, ranges, or error bars, an answer asserting an exact, specific figure beyond what the visual actually shows overstates the data's precision.

Do not claim more numerical precision than the figure provides.

Figure 1 reports the tree-canopy-only group's average reduction as 6.2°F with a range extending from roughly 4.8°F to 7.6°F across neighborhoods; a question asking for the "typical" reduction in this group is best answered using the reported average, 6.2°F, rather than by asserting that every neighborhood in the group experienced exactly that reduction, or by treating the endpoints of the range as if they were the typical value.

Respecting the visual's own precision, an average with a stated range of neighborhood-level variation, rather than reading more exactness into it than it actually offers, prevents both understating and overstating what the data supports.

When text and visual seem to disagree, identify exactly what each one measures. When a passage's claim and a visual's data appear to conflict, the first step is not to decide which source is more trustworthy, but to check precisely what each one measures: the same variable, the same population, the same time window, and the same scope. Many apparent contradictions between text and visual resolve once this precise comparison reveals that the two sources were never actually measuring the same thing in the first place.

Match the scope of the text claim to the scope of the visual.

If a hypothetical passage claimed that "canopy expansion produces the most dramatic results within the first year," while Figure 1's data is explicitly drawn from an average across the full 2021–2024 period, the apparent tension between a "first year" claim and a four-year average is not necessarily a genuine contradiction requiring one source to be wrong; it may simply mean the figure, as displayed, cannot confirm or deny a first-year-specific claim at all, since it measures something at a different time scope than the claim addresses.

Identifying this scope mismatch, rather than assuming the text and visual are making incompatible claims about the same thing, is the reliable way to resolve an apparent text-visual disagreement.

Combine claims, texts, and visuals without mixing their evidence.

A full Integration of Knowledge and Ideas question set draws on every mechanism in this chapter at once: identifying claims and their owners, weighing evidence for relevance and sufficiency, tracking how an argument handles a counterclaim, comparing two passages only after each is separately clear, and connecting a visual to the exact claim it bears on. The discipline that carries across all of it is the same discipline this chapter opened with: name the specific claim, the specific evidence, and the specific source before deciding how they relate to each other.

Compare literary passages on one specific dimension at a time.

Paired literary synthesis is not two separate comprehension exercises followed by one comparison question. The student must first preserve each passage as its own system, then identify a meaningful comparison dimension, then integrate evidence from both texts without forcing them into false symmetry. The hardest questions often ask for a conclusion that neither passage states alone but that becomes supportable once details from both are considered together.

Build two independent readings before comparing. Read Passage A for its own character relationships, structure, turning point, and theme. Do the same for Passage B. Do not begin by searching for a similarity. A premature comparison encourages vague labels such as "both are about family" that erase the specific work each text is doing.

Compare on one dimension at a time. A useful comparison names the dimension first: response to loss, treatment of responsibility, narrator distance, use of an object, structure of the ending, or another specific feature. "Both passages are similar" is not a conclusion. "Both use an ordinary object to make a character confront a relationship, but one object carries inherited obligation while the other exposes avoidance" is a comparison.

Distinguish shared situation from shared meaning. Two passages may place characters in similar situations and still develop different ideas. A character who returns an inherited object and a character who keeps one may both be responding to grief, but the opposite actions can express different understandings of what remembering requires.

Track how each character changes, not only where each begins. A cross-text character question often depends on movement. Compare initial belief to final belief in each passage. Two characters may start from opposite positions and move toward similar conclusions, or start similarly and diverge.

Compare narrator distance and knowledge. A first-person retrospective narrator can interpret earlier behavior directly. A third-person limited narrator may reveal a character more indirectly through action, perception, and withheld thought. Differences in narrative access affect what each passage can state openly and what must be inferred.

Mini-drill. If Passage A's adult first-person narrator says, "I know now why I lied," while Passage B's third-person narration only shows a character hiding a letter and avoiding a question, Passage A gives explicit retrospective interpretation; Passage B requires inference from behavior. A synthesis question about motive must respect that difference in certainty.

Compare the job of a repeated object or image. An object can function differently across passages even when both objects are inherited, damaged, repaired, or repeatedly handled. Ask what the object makes visible about the character, what changes in its meaning, and whether the passage treats keeping, repairing, discarding, or returning it as significant.

Mini-drill. Passage A: a character repairs a damaged watch because refusing to let it fail has become a way of refusing to accept a parent's death. Passage B: a character repairs a damaged chair so a younger sibling can continue using it. Both repair an inherited object. The comparison is not "both value tradition." In A, repair functions as resistance to loss; in B, repair functions as practical continuity.

Use structure as evidence for theme. A passage that ends with a decisive action may frame change as a choice. A passage that ends with an unresolved observation may frame change as recognition without resolution. When comparing themes, the ending matters because it shows what the text permits the character to settle and what remains unsettled.

Mini-drill. If one passage ends with an action and the other with a question, both may involve recognition, but only the first clearly resolves what the character will do. A choice claiming that both characters reach complete resolution would overstate the second passage.

A synthesis must be true of both passages. A tempting choice may describe Passage A perfectly and Passage B only loosely. That is not synthesis. Test each clause of the answer against both passages separately before accepting the combined statement.

Do not confuse contrast with contradiction. Two texts can emphasize different aspects of the same larger idea without opposing each other. If one passage shows that preserving an object can become avoidance and another shows that preserving an object can become care, the passages do not necessarily disagree about preservation. They may instead show that the meaning of keeping something depends on why and how it is kept.

The strongest cross-text inference often uses a shared principle. After identifying the differences, ask what broader principle can explain both. The correct synthesis is usually narrower than a universal moral but broader than a summary of either individual plot.

Mini-drill. Passage A ends with a character giving away a coat that belonged to a deceased relative. Passage B ends with a character deciding to keep a handwritten recipe. The actions differ, but both can support the broader idea that remembrance depends on choosing what relationship to have with the past, not automatically keeping or discarding every object connected to it.

Practice with passages you have already studied.

Answer each question using Passage E, Passage F, and Figure 1 as indicated. Locate the specific textual or graphical evidence that supports your choice before selecting it.

Questions 1–10: Passage E only