LSAT Inference Questions

Inference questions ask you to draw a conclusion from the statements in the stimulus. The stem may say that an answer is “most strongly supported,” “must be true,” or can be properly inferred. Read the exact wording: some questions require certainty, while others ask for the best-supported choice. In either case, the answer must stay within the evidence provided.

Reason from the given facts

Unlike a strengthen question, an inference question does not ask you to add new evidence. Unlike a necessary assumption question, it does not ask what an argument depends on. Your task is to identify what the premises already establish. Avoid outside knowledge, even when the topic seems familiar.

Example: “Every branch that opens on Sunday has a security guard. Some branches open on Sunday.” You can infer that some branches have a security guard. You cannot infer that every branch has a guard, because only some open Sunday. You also cannot infer that every guarded branch opens Sunday because the conditional runs one way.

Translate the stimulus into short statements before considering choices. The point is not to symbolize everything, but to make relationships visible. If a stimulus says one group is included in another, preserve that inclusion. If two facts refer to different populations, do not combine them as though they refer to the same people.

Preserve quantifiers

Quantifiers tell you how broad a claim is. “All” applies to every member of the group. “Some” means at least one. “Most” means more than half, but does not identify a particular member. “Many” and “few” are not exact percentages. “Not all” means there is at least one exception, while “none” rules out every member.

Suppose: “Most employees who completed the workshop submitted a project. Some employees who submitted a project received awards.” You cannot infer that any workshop participant received an award. The “some” award recipients may be outside the workshop group. The two statements share a phrase, but not enough information to connect the individuals.

Suppose instead: “All workshop participants submitted a project. Some workshop participants received awards.” Now you can infer that some project submitters received awards, because the award recipients are within the workshop group and all workshop participants submitted projects. The small change in the quantified relationship creates a valid chain.

Combine conditional rules carefully

For a conditional statement A → B, A guarantees B. Its contrapositive not B → not A is equivalent. If A → B and B → C, then A → C. These tools support valid inferences, but the converse B → A and inverse not A → not B do not follow.

Original example: “Every shipment classified as refrigerated requires a temperature log. Every shipment with a temperature log is reviewed by a supervisor. Shipment K is classified as refrigerated.” You can infer that K has a temperature log and is reviewed. If another statement says K was not reviewed, then K cannot be refrigerated under those rules. But a log alone does not prove refrigeration; other shipments might also require logs.

When a question includes multiple rules, link them only if the middle term matches. “Every archived file has a date stamp” and “every file with a date stamp has an owner” support archived → stamped → owner. But “some files have date stamps” and “some stamped files have owners” do not establish that a particular file is both archived and owned.

Statements about groups and overlap

A frequent inference challenge is determining whether two groups overlap. “Some researchers are grant recipients” and “some grant recipients work remotely” do not prove that a researcher works remotely. The two “some” statements may refer to different grant recipients. A shared category does not guarantee that the same individuals are involved.

A simple diagram can help. Draw circles for the groups and mark only the overlap the stimulus guarantees. If the facts do not require two groups to intersect, do not assume that they do. This is especially useful when the choices offer a broad claim about a population based on separate facts about its subgroups.

“Most A are B” and “most A are C” do imply that some A are both B and C, because each group contains more than half of A. For example, if 60 percent of a class completed a project and 70 percent attended a seminar, at least 30 percent must be in both groups. On an LSAT question, however, do not invent exact group sizes or percentages unless the stimulus gives enough information. The general overlap principle is what matters.

Must be true versus most strongly supported

A “must be true” stem requires an answer that follows with certainty from the premises. If an answer could be false while all the premises remain true, eliminate it. “Most strongly supported” may allow a conclusion that is highly supported but not strictly entailed, depending on the stimulus. Still, it is not permission to guess what is probably true in the real world.

When the wording is “most strongly supported,” compare the degree of support across choices. A modest claim closely tied to the facts often beats an ambitious explanation. If the stimulus says that a program's participants had higher scores than nonparticipants, it may support that participants scored higher in the sample. It does not by itself establish that the program caused the difference or that all future participants will do better.

A worked inference set

Facts: “All permit applications submitted after Friday are reviewed the following week. No application missing a required signature is approved. Some applications reviewed the following week are approved.” Which statement must be true? Any application submitted after Friday is reviewed the following week. Any approved application has all required signatures. Some applications reviewed the following week are approved. The first and second follow from the rules, and the third is directly stated.

What does not follow? You cannot infer that any application submitted after Friday is approved. The first rule only tells you when it will be reviewed. You cannot infer that every application reviewed the following week is approved; the statement says only some are. You also cannot infer that all approved applications were submitted after Friday, because the rules do not exclude earlier applications.

Now add: “All approved applications are published.” From the statement that some applications reviewed the following week are approved, you can infer that some applications reviewed the following week are published. You still cannot infer that every published application was reviewed the following week; publication may occur through another process. This is a chain that preserves the quantifier “some” and the direction of the rules.

Eliminate attractive but unsupported choices

A wrong choice may be true in ordinary life but not established by the stimulus. An argument about a school's admissions process does not let you assume every applicant has a recommendation letter. A passage about a medical study does not let you assume the treatment is safe just because no side effect is mentioned. Use only what is stated and validly implied.

Other distractors strengthen a premise, reverse a conditional, broaden a quantifier, or attach facts about one group to another. If an answer says “all” when the premise says “some,” check whether the extra strength is justified. If it uses the same vocabulary as the stimulus, confirm that the relationship is actually preserved.

Some choices merely repeat a fact. Repetition can be correct if the stem asks what follows, but it may not be the strongest answer if the other choice combines two facts into a new valid inference. Compare each choice with the complete set of premises, not just one sentence.

A method for tough stimuli

1. [object Object]
2. [object Object]
3. [object Object]
4. [object Object]
5. [object Object]
6. [object Object]

This process is deliberately conservative. Inference questions reward restraint. A choice that says less but follows from the facts is better than an interesting theory that needs one extra assumption. Your goal is not to explain why the facts occurred; it is to determine what the facts establish.

Mini practice with quantifiers

Facts: “Most volunteers completed orientation. All volunteers who completed orientation received a badge. Some badge recipients worked the evening shift.” What follows? You can infer that most volunteers received badges, because every volunteer in the majority who completed orientation receives one. You cannot infer that most volunteers worked evenings; the evening-shift statement concerns some badge recipients, who could include people outside the volunteer group.

Choice A: “All badge recipients were volunteers.” This reverses the membership direction and does not follow. Choice B: “Some volunteers worked evenings.” The evening workers may not be volunteers. Choice C: “Most volunteers received badges.” This follows from most volunteers completing orientation and the universal badge rule. Choice D: “Every volunteer completed orientation.” It changes most to every. Choice C is the supported answer.

This example uses a majority with a universal conditional. If more than half of volunteers completed orientation, and every person who completed it got a badge, then more than half got badges. The conclusion does not depend on the exact group size. But the separate fact about some badge recipients on the evening shift does not identify their membership in the volunteer group.

Review practice effectively

For every missed inference question, identify the exact extra claim you added. Did you assume two subgroups overlapped, reverse a rule, treat “most” as “all,” or turn a correlation into a cause? Then solve a fresh question with that pattern. If you guessed correctly, explain why the answer follows so that the result reflects a method rather than luck.

When an explanation says an option is “too strong,” specify the word that makes it too strong: all, always, only, must, or never. When it says “unsupported,” name the missing connection. The more precisely you diagnose an error, the easier it is to prevent it on a new question.

Sources

• [object Object]
• [object Object]