Reading Comprehension
What the section tests
Reading Comprehension (RC) is 50 multiple-choice items in 60 minutes based on three long passages, typically about 1,500 words each and often drawn from scientific or social-scientific writing. Questions ask for the main idea, explicitly stated details, inferences, the author's tone or purpose, vocabulary in context, and application of a passage's principle to a new case. The passages are dense, so the section tests navigation and evidence discipline as much as reading speed.
Core rules and concepts
- Read for structure first: identify the thesis, each paragraph's job, and any shift in argument (however, yet, by contrast).
- Distinguish stated from implied. An inference must follow from the text, not from your own science background.
- For detail questions, return to the passage and match the exact wording; the correct answer often paraphrases a single sentence.
- For main-idea questions, reject choices that are true but too narrow, too broad, or only one paragraph's point.
- For tone and purpose, watch evaluative adjectives and verbs: criticizes, qualifies, and documents are different claims.
- For EXCEPT and least-likely items, mark three choices supported by the passage and choose the remaining one.
Worked example
A passage states: After the treatment, nitrogen uptake rose sharply, but the authors caution that temperature also increased during the same period, so the cause of the increase remains uncertain. Question: Which is best supported? A likely correct answer is that the study cannot isolate the treatment as the sole cause of increased uptake. Reasoning: the passage reports a correlation and explicitly flags a confound. A choice saying the treatment caused the increase overstates the evidence; a choice saying temperature had no effect is unsupported. The correct answer tracks the author's caution, not the raw result.
Traps that cost marks
- Choosing an answer because it is scientifically true rather than because the passage supports it.
- Confusing the author's view with a researcher's view quoted in the passage.
- Missing a reversal word such as not or except in the question stem.
- Spending too long on one detail question and leaving later passages unread.
- Selecting an extreme answer (always, never, proves) when the passage uses qualified language (may, suggests).