LSAT Inference Questions: What You Can Actually Prove

Inference (or "must be true") questions ask which answer is guaranteed by the statements in the stimulus. You are not finding the main point and you are not strengthening anything. You are looking for the claim that has to be true if everything you were told is true.

The mental shift: in most LR questions the stimulus is an argument with a gap. In inference questions, treat the stimulus as a set of facts and ask what follows necessarily.

Step 1: Read the stimulus as a list of given facts, noting any conditional statements and quantifiers ("all," "some," "most," "none").

Step 2: Combine statements where they connect — chained conditionals and overlapping groups are the usual source of the right answer.

Step 3: Pick the answer that cannot be false given the facts. If you can imagine the facts being true while the answer is false, eliminate it.

"Must be true" demands airtight proof; the answer follows with certainty. "Most strongly supported" allows the best-supported answer even if it is not ironclad. The distinction changes how strict you should be: on must-be-true, reject anything that merely seems likely.

Too strong: an answer that says "all" when the stimulus only supports "some."

Out of scope: an answer that introduces a new term the stimulus never mentioned.

Could-be-true: an answer that is consistent with the facts but not required by them. On must-be-true, "could be true" is wrong.

Verbloom drills the combine-the-conditionals step so the provable answer is the one you build, not the one that sounds nice.