Start here: the section asks the same few things
Strong Reading and Writing performance depends on a small set of recurring decisions. The subject can change from ecology to legal history to literary criticism, yet the question still asks for one of the same things: the precise word, the strongest supported inference, the evidence that matches a claim, the exact point of agreement between two texts, the grammatical boundary, the logical transition, or the notes that satisfy a stated goal.
This opening unit walks through each of those decisions once, with a handful of worked questions per decision. As you work, notice that several choices often survive a quick reading. One word, one relationship, one clause boundary, or one evidentiary limit usually decides the question. The chapters that follow develop every decision in full, including harder variants and close-answer cases.
Words in Context: precision, force, and direction
Hard Words in Context questions often place several plausible academic words in the same answer set. The sentence fixes three things: the force of the word (how strong), its direction (toward or against the idea), and the relationship it must express.
Consider a historian who argues that newly discovered letters do not overturn the prevailing account of a negotiation but do ______ its scope by showing that informal intermediaries mattered. The letters narrow the reach of the existing account. They do not reject it (repudiate), make it worse (exacerbate), or prove it genuine (authenticate). The needed word is circumscribe: limit or define the bounds of something.
The same habit solves the others. If an apparent decline disappears once researchers correct for sampling depth, the earlier pattern was anomalous, meaning inconsistent with the expected pattern. Pervasive concerns extent, salient concerns prominence, and intractable concerns resistance to solution. A narrator who repeatedly gestures toward reform while avoiding commitment to any program makes the novel's stance equivocal: ambiguous in position. A revised procedure that does not eliminate error but shrinks its effect attenuates it: reduces its force without removing it.
Notice what a useful vocabulary record looks like. It keeps the tested distinction, not only the definition:
- circumscribe = limit the scope
- equivocal = ambiguous in position
- attenuate = reduce the force
- anomalous = inconsistent with the expected pattern
That form of memory transfers to new sentences. The next unit organizes the book's vocabulary into groups built around exactly these distinctions.
Inference: control the strength of the conclusion
Hard inference questions can leave two or more choices sounding reasonable on a quick read. The passage supports one of them at exactly the strength stated. Your job is to match the strength of the conclusion to the strength of the evidence, neither stronger nor weaker.
Take a study in which moths exposed to broad-spectrum artificial light began feeding later than moths exposed to amber light, while total feeding time did not differ. The study supports a claim about timing: the spectral composition of light can affect when feeding begins. It does not support a claim about total food intake (not measured), about amber light having no effect (it clearly did something), or about survival (never tested).
Design matters too. If speakers of two dialects outperform single-dialect speakers on a vowel contrast, but the groups also differ in age and region, the difference is real but confounded. The strongest defensible claim is association, not causation. And when an experiment shows stomata closing normally in stressed plants whose hormone H did not rise, the result suggests only that closure can occur without the predicted rise. It does not identify the real mechanism, and it does not generalize to every drought response.
Words that often decide these questions include can, may, tends to, is associated with, is consistent with, supports, requires, demonstrates, and causes. Read them as claims of different strength.
Command of Evidence: the evidence must perform the requested job
Evidence questions are not solved by finding a sentence on the same topic. The evidence has to support the claim as written.
If a scholar claims that a playwright uses interruptions to show that a character's authority is less secure than it appears, the quotation that works is the one in which the character begins a decree twice and is cut off both times. Quotations that only say the hall fell silent, or that he served for a decade, or that the decree concerned an old dispute, establish context without demonstrating the claim.
Scientific claims work the same way. If a biologist proposes that a seed coating improves survival specifically by reducing water loss, the strongest finding connects the coating to both the intermediate process and the outcome in the relevant condition: under dry conditions, coated seeds lose water more slowly and survive longer. A finding that coated seeds survive better in both dry and humid conditions does not isolate the proposed mechanism.
Quantitative Evidence: identify the comparison before reading the values
A difficult graph or table question may contain more data than the claim requires. The central task is identifying the comparison embedded in the sentence. Before touching the data, name the required quantity: endpoint, difference, absolute change, percentage change, rate, proportion, rank, or association.
If a researcher claims the largest absolute increase occurred in population C, then post-intervention rank and percentage change answer different questions. The relevant fact is the size of each increase (19 for C against 16 and 13). In a table, the species with the higher endpoint is not automatically the species with the larger within-species change. And a correlation of −0.62 between canopy density and understory temperature justifies only a statement about direction and strength of association: denser canopies tend to go with lower temperatures. It does not justify a percentage effect, a claim that density explains 62 percent of the difference, or a reversed causal arrow.
Cross-Text Connections: isolate the exact proposition
When two texts respond to each other, locate what both accept before you locate where they split. If Text 1 attributes a bird decline to habitat loss and Text 2 accepts the habitat loss but adds an introduced predator, both authors agree that habitat loss contributed. They disagree about whether it was sufficient. Likewise, if Text 2 doubts a trade interpretation of artifacts and asks for chemical sourcing, it accepts the observations and questions the inference drawn from them.
Cross-text questions often separate evidence from interpretation, contribution from sufficiency, and agreement on a result from agreement on its explanation. Write the shared proposition in one short sentence before checking the choices.
Text Structure and Purpose: ask what the sentence contributes
Do not ask what a sentence says; ask what it does in the passage. A sentence introducing genetic evidence that conflicts with a central assumption of an old explanation limits that explanation's power. It is not background, not a strengthening example, and not a complete replacement theory. Similarly, a quotation praising a city's “perfect order” placed right after delays, contradictory instructions, and misplaced records creates irony. The surrounding details give the quotation its function; the words alone would not reveal it.
Standard English Conventions: locate the structure inside the sentence
Dense grammar questions become manageable once the sentence is reduced to its core clauses. The difficulty comes from the material placed between those clauses.
- Subject-verb agreement: in “The collection of field notes from three expeditions, together with several later annotations, ______ the basis of the catalog,” the subject is collection, which is singular. The plural nouns in between do not control the verb.
- Opening modifiers: after “After comparing the digitized maps with the original notebooks,” the next word group must name whoever did the comparing. “The research team” can be the subject; “inconsistencies” cannot.
- Parenthetical interrupters: in “later handwriting evidence, however, showed,” the word however sits inside a single independent clause, so it needs a comma on each side. A semicolon there would require the words before it to form a complete clause, and they do not.
- Verb tense: damage that occurred before another completed past event (the 2019 examination) takes the past perfect: had damaged.
Transitions: solve the relationship first
Decide how the second sentence relates to the first before looking at the choices. A later result that differs from what an early similarity might lead you to expect calls for a contrast such as nevertheless. A sentence that gives a limitation to a success just described calls for however. A sentence that narrows a broad statement to one subset calls for specifically. Match the relationship, then the word.
Rhetorical Synthesis: satisfy the goal and verify the facts
These questions become harder when two choices appear to satisfy the goal. Use the goal to identify the required information, compare the choices, then verify the survivor against the notes. If the goal is to emphasize a result that was consistent across locations, the answer states the result shared by all locations. If the goal is to present a critic's interpretation, the answer must include both the critic's claim and what the imagery is said to do. In harder items, two choices can fit the goal while only one preserves the notes accurately.
Finding the exact point of disagreement
At higher scores, two choices often share most of their content. Compare the words where they separate: causes versus is associated with; shared explanation versus shared observation; sentence content versus sentence function; endpoint versus change; accurate attribution versus a shifted attribution. Locate the exact difference, return to the relevant line or datum, and decide that point.
What to build from practice
The College Board organizes Reading and Writing into four official content domains: Information and Ideas, Craft and Structure, Expression of Ideas, and Standard English Conventions. Within those domains, this course teaches the finer-grained skills you actually need to practice: Central Ideas and Details, Command of Evidence, Inferences, Words in Context, Text Structure and Purpose, Cross-Text Connections, Rhetorical Synthesis, Transitions, Boundaries, and Form, Structure, and Sense. Your practice can be concentrated on the exact family producing your misses.
Keep an error record built around the deciding feature, such as:
- “chose causation from associative evidence”
- “read equivocal as negative instead of ambiguous”
- “answered final level when the claim asked for change in the gap”
- “treated shared evidence as shared interpretation”
- “attached an opening modifier to the wrong subject”
Precise labels make repeated errors visible. The subject matter may stay unfamiliar. Practice should make the underlying decision recognizable.