GMAT RC Inference Questions: The Evidence-Bound Approach
GMAT Reading Comprehension inference questions ask: 'Which of the following can be inferred from the passage?' or 'The passage implies which of the following?' The correct answer is a conclusion that follows directly and necessarily from specific text in the passage. Wrong answers are plausible extensions, real-world knowledge, or logical conclusions that require stepping beyond what the passage actually states.
The evidence-bound approach: every correct inference answer must be verifiable through a specific location in the passage. Before evaluating answer choices, ask: 'What would an inference from this passage look like?' It would be: (1) more specific than the passage's general claims, connecting two stated facts to reach a conclusion neither states explicitly; or (2) the logical consequence of a qualified claim in the passage ('if X is true and Y is true, then Z follows').
For each answer choice, apply the two-step test: (1) Identify the passage text that would support this inference. If you cannot find supporting text, eliminate. (2) Verify that the inference is directly supported — not just consistent with — the text. 'Consistent with' is a lower bar than 'directly supported'. GMAT wrong answers are typically consistent with the passage (they don't contradict it) but go beyond what the passage directly establishes.
The most common RC inference wrong answer types: (1) Real-world knowledge intrusion — an answer that is true in the real world but not stated or implied by this specific passage. (2) Too-strong inference — uses 'always', 'never', 'all', 'none' when the passage uses qualified language ('often', 'tends to', 'in some cases'). (3) Scope extension — the passage discusses a narrow case (studies of a specific species) and the inference draws a broad conclusion (all organisms of that type). (4) Opposite inference — draws the logically opposite conclusion from what the passage supports, exploiting misreads of negation or conditional language.
Practice protocol for RC inference: after reading each passage, before looking at questions, generate three plausible inferences from the passage content. Write them down. Then check whether any question tests exactly those inferences. This active synthesis practice builds the habit of reading for implication, not just information.