LSAT Logical Reasoning — Inference and Resolution (Deep-Dive Study Notes)

Must Be True

Must Be True (MBT) questions ask you to find a statement that is logically forced by the stimulus. If the stimulus is true, the correct answer cannot be false. This is the strictest kind of inference task on LSAT Logical Reasoning.

What it is

An MBT question is about deduction. You are not being asked what is likely, reasonable, or a good idea. You are being asked what follows with certainty from the given facts.

A helpful way to frame it: the stimulus gives you a small “world” of information. The correct answer is something that is true in every version of that world that satisfies the stimulus.

Why it matters

MBT questions train the core LR skill: separating what is stated (or strictly implied) from what is merely suggested. Many wrong answers are tempting because they sound plausible or align with common sense—but the LSAT rewards what can be proven from the text, not what seems reasonable.

This skill also connects to other LR tasks:

  • MBT is closely related to Cannot Be True (one is about necessary truth, the other about impossibility).
  • It also overlaps with Most Strongly Supported, but MBT is stricter—“supported” can be strong without being guaranteed.

How it works (a reliable process)

  1. Treat the stimulus as facts, not an argument (unless it clearly is one). Many MBT stimuli are just descriptions, rules, or sets of statements.
  2. Translate relationships into clean logic. You don’t need formal symbols, but you do need clarity about what depends on what.
    • Words like all, none, only, unless, except, most, some are high-impact.
    • Watch for conditional relationships (if/then), quantifiers (some/most/all), and exclusions.
  3. Look for what must hold in every case. Common sources of forced inferences:
    • Chaining conditionals: If A → B and B → C, then A → C.
    • Contrapositives: If A → B, then not-B → not-A.
    • Mutual exclusivity: If X cannot happen with Y, then X implies not-Y.
    • Quantifier implications: If all A are B, then anything that is A is also B.
  4. Use answers as hypotheses, not invitations. For each choice, ask: “Can I prove this from the stimulus?” If you can imagine a scenario where the stimulus is true and the answer is false, the answer is out.

A strong mindset shift: you are not picking the answer that best matches the stimulus’s “vibe.” You’re picking the answer that survives every possible counterexample.

Show it in action (worked examples)

Example 1

Stimulus:
All of the conference speakers are researchers. Some of the researchers are engineers.

Question: Which one of the following must be true?

Reasoning:

  • “All speakers are researchers” tells you: Speaker → Researcher.
  • “Some researchers are engineers” tells you: at least one Researcher is an Engineer.
  • Notice what you don’t know: whether any speakers are engineers.

Must-be-true inference:

  • At least one researcher exists (because “some researchers are engineers” guarantees at least one researcher).

Correct-style answer: “At least one researcher is an engineer.” (This is exactly restating the second premise.)

Tempting wrong-style answer: “Some speakers are engineers.” Not forced—could be that the engineer-researchers are not speakers.

Example 2

Stimulus:
If the museum is open, then the lights are on. The lights are not on.

Question: Which one of the following must be true?

Reasoning:

  • Open → Lights on.
  • Not (Lights on).
  • By contrapositive: Not (Lights on) → Not open.

Must-be-true inference: The museum is not open.

What goes wrong (common traps)

  • Answer choices that add new ideas. MBT answers rarely introduce a new concept that wasn’t already in the stimulus.
  • Confusing “could be true” with “must be true.” If an answer merely fits, it’s not enough.
  • Mistaking everyday causality for logical force. Even if something seems like it “should” follow in real life, you need it to follow from the text.
Exam Focus
  • Typical question patterns:
    • “Which of the following can be properly inferred?”
    • “The statements above, if true, most strongly support which of the following?” (Sometimes this is actually MSS—read carefully.)
    • “Which of the following must also be true?”
  • Common mistakes:
    • Picking an answer that is strongly suggested but not guaranteed—avoid by actively trying to imagine a counterexample.
    • Forgetting the contrapositive for conditional statements—practice flipping and negating correctly.
    • Over-interpreting quantifiers (e.g., treating “some” like “most” or “all”).

Most Strongly Supported

Most Strongly Supported (MSS) questions ask for the answer choice that is best backed by the stimulus, even if it is not 100% guaranteed. Think of MSS as the “best inference” rather than the “airtight deduction.”

What it is

In MSS, the correct answer is the statement that has the greatest support relative to the other options. It may still be possible (in a far-fetched way) for it to be false while the stimulus is true—but it should be harder to deny than any competitor.

A useful analogy:

  • MBT is a mathematical proof.
  • MSS is a courtroom standard where you’re picking the conclusion most supported by the evidence on the record.

Why it matters

MSS questions test whether you can:

  • Recognize what the stimulus makes probable versus certain.
  • Compare answers strategically—because the correct answer wins by being most supported, not necessarily “perfect.”

This also connects to Resolve the Paradox: both ask you to work with what’s given without adding unsupported assumptions. MSS just aims at the strongest conclusion; Resolve aims at the best reconciliation.

How it works (a reliable process)

  1. Extract the key facts and any limitations. MSS often includes statistics, surveys, or patterns—note what’s measured and what isn’t.
  2. Predict the direction of a supported inference. You may not predict the exact wording, but you can predict what kind of claim would be reasonable.
  3. Rank answers by support. For each answer, ask:
    • Does it restate or lightly repackage the stimulus (often strong)?
    • Does it make a modest extension that fits the facts (possibly correct)?
    • Does it go beyond the evidence (often wrong)?
  4. Prefer weaker, more careful language when support is limited. MSS correct answers often avoid absolute claims like “all,” “never,” or “must” unless the stimulus really warrants them.

Show it in action (worked examples)

Example 1

Stimulus:
In a study of 1,000 commuters, those who listened to podcasts during their commute reported higher satisfaction with their commute than those who did not.

Question: Which of the following is most strongly supported?

Reasoning:

  • We have an association: podcast listeners reported higher satisfaction.
  • We do not have causation (podcasts might not be the reason).

Most supported-style answer: “Among the commuters studied, podcast listeners tended to report greater commute satisfaction than non-listeners.”

  • This sticks closely to what was measured.

Common tempting wrong answers:

  • “Listening to podcasts causes commuters to be more satisfied.” (Adds causation.)
  • “Most commuters listen to podcasts.” (No frequency information.)
Example 2

Stimulus:
Every employee who received an outstanding performance rating last year completed the new training course. Some employees who completed the new training course did not receive an outstanding rating.

Question: Which statement is most strongly supported?

Reasoning:

  • Outstanding → Trained.
  • Some Trained → Not Outstanding.
  • A safe inference: at least one trained employee is not outstanding (explicit).
  • Another strong inference: not everyone trained is outstanding.

Correct-style answer: “Not all employees who completed the training course received an outstanding rating.”

  • This is a clean rephrase of “some trained did not receive outstanding.”

What goes wrong (common traps)

  • Overchoosing certainty. Students often import MBT standards and reject the right answer because it’s not perfectly guaranteed. In MSS, perfection is not required—comparative strength is.
  • Ignoring scope and measurement. If the stimulus is about self-reports, the conclusion must be about self-reports, not necessarily objective reality.
  • Falling for extreme wording. If the stimulus doesn’t justify “all/none/always,” those are usually too strong.
Exam Focus
  • Typical question patterns:
    • “Which of the following is most strongly supported by the information above?”
    • “The statements above most strongly support which of the following?”
    • “Which conclusion can be most reasonably drawn?”
  • Common mistakes:
    • Treating correlation as causation—avoid by asking, “Did they claim a cause, or only a relationship?”
    • Choosing an answer that’s true in real life but not grounded in the stimulus—stay inside the text.
    • Missing that a modestly worded answer can beat a flashier but overreaching one.

Cannot Be True

Cannot Be True (CBT) questions ask you to identify the answer choice that is impossible if the stimulus is true. In other words, the correct answer must be false given the facts.

What it is

CBT is the “negative” mirror of MBT:

  • MBT: true in all allowable scenarios.
  • CBT: true in no allowable scenarios.

Many CBT questions are built on:

  • Rule sets (like mini logic games in LR form)
  • Conditional constraints
  • Mutually exclusive categories

Why it matters

CBT tests whether you can use constraints to eliminate possibilities cleanly. It’s also a strong diagnostic of careful reading—one missed word like “only if” or “unless” can flip what’s possible.

CBT also teaches a powerful general LR habit: when you want to test a claim, try to fit it into the facts. If it breaks the system, it’s out.

How it works (two practical methods)

Method A: Build the “possible world” boundaries
  1. List the stimulus constraints.
  2. Combine them (chain conditionals, apply exclusions).
  3. For each answer, ask: “Can I create at least one scenario where this could happen without violating any rule?”
  • If yes, it is not the CBT answer.
  • If no—every attempt breaks a rule—it’s the CBT answer.
Method B: Use contradiction hunting

Instead of building a full scenario, try to see if the answer directly contradicts something forced by the stimulus.

  • If you can derive an MBT inference, any answer that denies it is automatically CBT.

Show it in action (worked examples)

Example 1 (conditional contradiction)

Stimulus:
If Carla works on Saturday, then Devin works on Saturday. Carla works on Saturday.

Question: Which of the following cannot be true?

Reasoning:

  • Carla Saturday → Devin Saturday.
  • Carla Saturday is true, so Devin Saturday must be true.

Cannot-be-true answer: “Devin does not work on Saturday.”

  • This contradicts what must follow.
Example 2 (rules/exclusions)

Stimulus:
A library will host exactly two events next week: a lecture, a workshop, and a film screening are the only possibilities. The library will not host both a lecture and a film screening.

Question: Which option cannot be true?

Reasoning:

  • Exactly two events.
  • Choices are {Lecture, Workshop, Film}.
  • Not (Lecture and Film together).

Possible pairs:

  • Lecture + Workshop (allowed)
  • Workshop + Film (allowed)
  • Lecture + Film (forbidden)

Cannot-be-true answer: “The library will host a lecture and a film screening.”

What goes wrong (common traps)

  • Confusing “cannot be true” with “not supported.” An answer can be unsupported yet still possible.
  • Failing to use the “exactly/at least/at most” numbers. These words create hard constraints.
  • Missing that CBT requires impossibility. If you can build even one consistent scenario, it’s not the answer.
Exam Focus
  • Typical question patterns:
    • “Which of the following cannot be true?”
    • “Which scenario is inconsistent with the statements above?”
    • “Which of the following must be false?”
  • Common mistakes:
    • Eliminating an answer because it seems unlikely or odd—odd is allowed; only contradictions are forbidden.
    • Overlooking a single constraint while checking choices—use a quick checklist of rules as you test each option.
    • Negating the wrong thing in a conditional—remember: the contrapositive flips and negates both sides.

Resolve the Paradox

Resolve the Paradox questions ask you to choose the answer that best explains how two (or more) seemingly conflicting facts can both be true. The stimulus presents a surprising outcome, discrepancy, or tension; your job is to remove the “mystery,” not to prove someone right.

What it is

A paradox here is not a formal logical contradiction like “A and not-A.” It’s usually an unexpected coexistence:

  • A plan that should work but doesn’t
  • A policy intended to reduce something but it increased
  • Two studies that seem to disagree
  • A trend that violates a general expectation

The correct answer supplies a missing piece that makes the situation coherent.

Why it matters

Resolve questions build real analytical reasoning: in the real world, data often clashes with expectations. The LSAT tests whether you can:

  • Identify the expectation driving the surprise
  • Find a factor that breaks that expectation (or shows it doesn’t apply)

These questions are also closely related to Strengthen/Weaken:

  • Strengthen supports a conclusion.
  • Resolve supports the coexistence of facts—often without any “conclusion” at all.

How it works (step-by-step)

  1. Spot the two sides of the discrepancy. Write them in your own words.
    • Fact A: what happened / what is observed.
    • Fact B: what you’d expect instead, or the conflicting observation.
  2. Identify the hidden assumption that creates the surprise. Usually it’s something like:
    • “All else stayed the same.”
    • “The measure captures what we care about.”
    • “The intervention affects the main driver.”
  3. Know what a good resolution looks like. A correct answer typically does one of these:
    • Shows the cases aren’t comparable (different populations, definitions, timing).
    • Introduces a new factor that counteracts the expected effect.
    • Clarifies that the apparent change is due to measurement, reporting, or selection effects.
  4. Avoid answers that take sides. You’re not choosing “which fact is wrong.” You’re choosing what allows both to stand.

A practical test: after you read an answer, you should be able to say, “Ah, then it makes sense that both statements could be true.”

Show it in action (worked examples)

Example 1 (counteracting factor)

Stimulus:
A city installed more bike lanes to reduce traffic congestion. Yet after the bike lanes were installed, congestion increased.

Question: Which of the following, if true, most helps resolve the apparent paradox?

Reasoning:

  • Expectation: more bike lanes → fewer cars → less congestion.
  • Observation: congestion increased.
  • Resolution needs a factor that explains the increase without denying the bike lanes were installed.

Resolution-style answer: “During the same period, a major bridge closed for repairs, forcing many drivers onto the city’s main roads.”

  • Now congestion can rise even if bike lanes help, because a bigger counter-force worsened traffic.

Wrong-direction answers (common):

  • “Many residents support bike lanes.” (Doesn’t explain congestion.)
  • “Some drivers dislike bike lanes.” (Attitude doesn’t resolve outcome.)
Example 2 (measurement/definition shift)

Stimulus:
A restaurant changed its recipes to reduce sodium. Laboratory tests show the new menu items contain less sodium than before. Yet customer surveys report that the food tastes saltier.

Reasoning:

  • Fact: sodium decreased.
  • Fact: perceived saltiness increased.
  • Good resolutions often involve perception, other ingredients, or comparison effects.

Resolution-style answer: “The new recipes use more acidic ingredients, which can make salty flavors more noticeable even when sodium levels are lower.”

  • Both can be true: less sodium but stronger perception of saltiness.

What goes wrong (common traps)

  • Trying to “solve” it by rejecting a fact. Many wrong answers imply one of the observations is mistaken, but the question typically treats them as true.
  • Choosing an answer that strengthens one side only. If it just explains why sodium dropped, that doesn’t explain why it tastes saltier.
  • Picking an answer that’s relevant but not reconciling. It’s easy to select something connected to the topic that doesn’t actually bridge the gap.

A useful mental template

When you see a discrepancy, ask:

  • “What would have to be true for this to make sense?”
    Common categories of resolutions:
  • Different baseline: comparing against a different reference point than you assumed.
  • Different group: the people measured aren’t the people you’re thinking of.
  • Offsetting cause: the fix helped, but something else hurt more.
  • Changed measurement: the metric moved even if the underlying reality didn’t (or vice versa).
Exam Focus
  • Typical question patterns:
    • “Which of the following, if true, most helps to resolve the apparent discrepancy?”
    • “Which of the following most helps to explain the surprising result?”
    • “The statements above are best reconciled by assuming which of the following?”
  • Common mistakes:
    • Treating it like Strengthen/Weaken of an argument—first identify the two facts that must be reconciled.
    • Choosing an answer that could be true but doesn’t connect the two sides—force yourself to articulate the bridge.
    • Over-assuming causation—sometimes the paradox is about measurement or comparisons, not causal impact.