2.3 Read an Academic Passage
Key Takeaways
- The current academic passage is about 200 words and is followed by five questions.
- Questions test stated information, inference, vocabulary in context, relationships among ideas, and purpose or organization.
- A compact claim-evidence map is more useful than detailed notes on a short passage.
- Answers must be supported by the passage; outside subject knowledge is not evidence.
Read an Academic Passage
The current task is deliberately compact. ETS describes an academic passage of about 200 words, followed by five questions. It is not the pre-2026 format built around 700-word textbook excerpts and ten questions. A passage may introduce a scientific process, explain a historical development, compare theories, or describe evidence for a claim. The subject can be unfamiliar, but the information needed to answer is in the text.
Build a rapid structure map
Read the whole passage before answering. At this length, a complete first read is faster and safer than searching sentence by sentence without context. Give each paragraph a short function label: definition, cause, example, contrast, evidence, or consequence. Then state the passage's central claim in one sentence.
Suppose a passage says that urban trees cool neighborhoods through shade and transpiration, then explains why benefits differ with canopy density. A useful map is: claim—two cooling mechanisms—condition affecting strength. Copying every temperature number would be less helpful. The map keeps the relationships visible when a question asks why the author mentions canopy density.
Five recurring reasoning demands
Stated information: locate the relevant sentence and match its meaning, not merely a repeated word. A distractor often borrows vocabulary while reversing the relationship or changing its scope.
Inference: combine facts that the text actually supplies. A valid inference is strongly supported even if it is not stated word for word. Reject choices that could be true in the real world but need an unstated assumption.
Vocabulary in context: substitute each option into the sentence, then check the surrounding claim. The everyday meaning may be wrong in the passage. For example, marked can mean noticeable rather than physically labeled.
Relationship or organization: identify what one sentence or paragraph does for another. It may give an example, qualify a generalization, supply a cause, or report competing evidence. State that function before looking at the choices.
Purpose: ask why the author included a detail. The best answer connects the detail to the local argument. “To provide information” is usually too vague; “to illustrate how water loss produces cooling” is testable against the passage.
Evidence-first answering
For every choice, point to a line or a relationship in the text. Use three filters:
- Accuracy: does the text support the claim?
- Scope: does the choice stay within the passage's limits?
- Logic: does it preserve cause, contrast, time, and certainty?
Watch qualifiers. Some is not all; may is not will; after does not automatically mean because of. Many attractive distractors become wrong through a single altered qualifier.
Manage the set
Use roughly the first quarter of the set time to read and map, then answer the five questions. This is a practice ratio, not an ETS-imposed per-task clock. If one question stalls, eliminate what you can, choose the best supported answer, and continue. The adaptive section clock is more important than perfecting one item.
Keep notes minimal: a three- to five-word main idea, paragraph functions, and perhaps one contrast arrow. Because the passage remains available while answering, extensive transcription wastes time. Revisit the exact sentence whenever two options remain.
Practice and review
Practice with passages near the official length. After each set, record the reasoning error rather than only the topic: missed qualifier, confused example with main claim, imported outside knowledge, or accepted a plausible but unsupported inference. Then rewrite the evidence in a short sentence. This turns a wrong answer into a reusable rule.
Do not train on obsolete question families as if they define the current test. Long-passage prose summaries and sentence-insertion routines can develop general reading, but they are not the blueprint for this task. Prioritize concise academic texts, five-question sets, accurate relationship tracing, and fast evidence checks.
The strongest routine is simple: read fully, map the argument, predict the answer in plain language, and test every option against the passage. Academic background may make a topic feel familiar, but textual evidence remains the scoring standard.
A two-choice audit
When two choices remain, compare them one phrase at a time. Mark the actor, action, object, condition, and certainty in each choice. Then align those elements with the passage. A choice can be mostly accurate yet fail because it assigns the action to a different group or omits a condition. If both choices appear supportable, reread the question stem: one may answer what happened while the other answers why it happened. This audit is faster with practice and prevents familiarity with a repeated noun from substituting for evidence.
What is the official current academic-passage configuration?
Which inference is safest?
Why label a paragraph as contrast, evidence, or example?