LSAT Logical Reasoning: Learning to Analyze and Evaluate Arguments

What an LSAT “Argument” Is (and How to Take It Apart)

Argument vs. information

In LSAT Logical Reasoning, most stimulus passages are trying to do one of two things:

• Present an argument: a set of statements where some statements (premises) are offered as support for another statement (conclusion).
• Present non-argumentative information: facts, descriptions, narratives, or competing viewpoints without a single claim being supported.

This distinction matters because most question types—strengthen, weaken, necessary assumption, flaw, etc.—only make sense if there is an argument to evaluate. If you misidentify an argument as mere information (or vice versa), you’ll apply the wrong toolkit.

The core parts: premises and conclusion

An argument is built from:

• Conclusion: the main claim the author is trying to get you to accept.
• Premises: reasons offered to support that conclusion.

A reliable way to find the conclusion is to ask: “If the author had to be wrong about just one thing for the whole argument to fail, what would it be?” That’s usually the conclusion.

Common conclusion and premise indicators

Indicator words are helpful, but not guaranteed.

• Conclusion indicators: “therefore,” “thus,” “hence,” “so,” “consequently,” “clearly,” “it follows that.”
• Premise indicators: “because,” “since,” “for,” “given that,” “after all.”

The LSAT often withholds these words or uses them misleadingly. So treat them as clues, not proof.

Subconclusions and argument structure

Many stimuli contain a subconclusion: a claim that is supported by earlier premises and then used as a premise to support the main conclusion. Think of it as a “stepping-stone” conclusion.

Why it matters: if you mistake a subconclusion for the main conclusion, you’ll evaluate the wrong support relationship and choose answers that address the wrong gap.

Background statements and context

A stimulus can include statements that are neither premise nor conclusion, such as:

• historical background
• definitions
• a concession (“While X is true, …”)
• framing (“Some people think …”)

These are often included to distract you into treating them as support.

Worked example: identifying parts

Stimulus: “Many people assume that exercise alone causes weight loss. But several studies show that people who exercise without changing diet lose little weight. Therefore, lasting weight loss requires dietary change.”

• Background/common belief: “Many people assume …”
• Premise: “Several studies show … lose little weight.”
• Conclusion: “Lasting weight loss requires dietary change.”

Notice the argument isn’t claiming exercise never helps; it’s making a necessity claim about diet.

Exam Focus

• Typical question patterns:
  • “The conclusion of the argument is…” (Main conclusion)
  • “Which statement is a premise?” / “The argument proceeds by…” (Method)
  • Flaw/assumption questions that require pinpointing what’s being supported
• Common mistakes:
  • Treating a strongly worded premise as the conclusion because it “feels central”
  • Missing a subconclusion and thinking the argument has a logical leap it doesn’t actually have
  • Using indicator words mechanically (e.g., assuming anything after “therefore” is always the main conclusion)

Understanding What the Question Stem Is Asking You To Do

Why the stem is half the problem

Logical Reasoning is not one skill; it’s a set of repeatable tasks. The question stem tells you which task to perform. Two questions can use the same stimulus but require opposite actions (strengthen vs. weaken). Your first job is to identify the job description.

Families of tasks

Most LR tasks fall into a few families:

1. Find something true (based on the stimulus)
   • Must Be True / Most Strongly Supported
2. Find what the argument needs
   • Necessary Assumption, Sufficient Assumption
3. Evaluate impact on the argument
   • Strengthen, Weaken
4. Explain a mismatch
   • Resolve/Explain a Paradox
5. Analyze reasoning form
   • Flaw, Method of Reasoning, Parallel Reasoning, Parallel Flaw
6. Extract or apply a general rule
   • Principle (justify), Principle (conform)
7. Clarify function
   • Role/Function of a statement

Knowing the family helps because it tells you what counts as a good answer. For example, a Must Be True answer must be forced by the text, while a Strengthen answer merely needs to make the conclusion more likely.

“Most strongly supported” is not “could be true”

One of the biggest mindset shifts is understanding the LSAT’s standard of proof. In “Most Strongly Supported,” you’re still looking for a statement that is strongly backed by the stimulus—not a speculative possibility. If several answers “could” happen, pick the one most anchored to what’s said.

“Except” and “least” stems

Some stems ask for the answer choice that does NOT do something (“All of the following strengthen EXCEPT…”). This flips the search: four answers will fit the job, one won’t.

A practical way to avoid errors is to mark in your mind that you are hunting for the odd one out, and to actively confirm the four that do match.

Exam Focus

• Typical question patterns:
  • “Which of the following, if assumed, enables the conclusion to be properly drawn?” (Sufficient Assumption)
  • “Which of the following would most weaken the argument?” (Weaken)
  • “Which of the following is most strongly supported by the statements above?” (MSS)
• Common mistakes:
  • Treating every question like an assumption question (trying to “fix” when you’re supposed to infer)
  • Forgetting the direction of impact (strengthen vs. weaken)
  • Missing “EXCEPT/LEAST” and choosing a strong supporter instead of the non-supporter

Conditional Reasoning: How “If…Then…” Logic Works on the LSAT

What conditional statements are

A conditional statement links a sufficient condition to a necessary condition:

• Sufficient condition: if this happens, it triggers the other.
• Necessary condition: this must be true whenever the sufficient condition is true.

In plain language: “If A, then B” means A guarantees B.

Common conditional indicators

LSAT conditionals often appear without the word “if.” Learn the common translations:

• “If A, then B”
• “A only if B” means: If A, then B (B is necessary)
• “A if B” means: If B, then A (B is sufficient)
• “A requires B” means: If A, then B
• “A is sufficient for B” means: If A, then B
• “A is necessary for B” means: If B, then A

The trap is reversing these, especially with “only if” and “unless.”

Contraposition (the valid inference)

From “If A, then B,” you can validly infer its contrapositive:

• If not B, then not A.

This matters because many LSAT arguments rely on reasoning that is only valid via contrapositive. If you don’t recognize it, valid logic can look like a leap.

What you cannot do: the classic fallacies

Two invalid moves show up constantly:

• Affirming the consequent: If A then B. B. Therefore A. (Invalid)
• Denying the antecedent: If A then B. Not A. Therefore not B. (Invalid)

In real life these feel tempting because B might be common when A happens, but the logic doesn’t guarantee it.

“Unless” statements

“Unless” can be converted into a conditional by treating it as “if not.”

• “A unless B” becomes: If not B, then A.

Often there’s also an equivalent alternative formulation: “If not A, then B.”

Example: “You won’t pass unless you study.”

• If you do not study, then you won’t pass.
• Equivalent: If you pass, then you studied.

Worked example: spotting the flaw

Stimulus: “If a product is truly organic, it will have certification. This product has certification. So it is truly organic.”

Form:

• If organic, then certified.
• Certified.
• Therefore organic.

That’s affirming the consequent. Certification could come from error, fraud, or a different standard.

Exam Focus

• Typical question patterns:
  • Flaw/parallel flaw questions built on conditional fallacies
  • Sufficient assumption questions where you supply the missing link (“If certified, then organic”)
  • Must Be True questions requiring contraposition
• Common mistakes:
  • Reversing “only if” (“A only if B” mistakenly read as “If B then A”)
  • Forgetting that “if” introduces the sufficient condition (not the necessary)
  • Treating contrapositive as optional rather than logically equivalent

Quantifiers and Scope: “Most,” “Some,” “All,” and “Not All”

Why quantifiers matter

A large share of LR arguments hinge on the difference between:

• Universal claims: “all,” “every,” “none,” “no,” “always”
• Existential claims: “some,” “at least one,” “many”
• Proportional claims: “most,” “usually,” “more than half”

The LSAT tests whether you notice when an argument illegitimately moves between these.

Common illegal shifts

A common flaw is moving from “some” to “all”:

• Premise: Some council members opposed the bill.
• Conclusion: The council opposed the bill.

Even “many” doesn’t justify “most,” and “most” doesn’t justify “all.”

Another common issue is switching the group:

• Premise about students at this school
• Conclusion about students in general

That’s a scope shift—the argument quietly changes who the claim is about.

Negating quantifiers (useful for assumptions)

Understanding negation helps with Necessary Assumption questions.

• Negation of “all” is “not all.”
• Negation of “some” is “none.”
• Negation of “most” is “50% or less.”

You don’t usually need to compute numbers; you just need to know what would contradict the claim.

Example: scope + quantifier

Stimulus: “Most residents who attend the town hall meeting oppose the new zoning plan. Therefore, most residents oppose the zoning plan.”

The premise is about a subgroup (attendees), not all residents. Even if most attendees oppose it, attendees might be unrepresentative.

Exam Focus

• Typical question patterns:
  • Flaw questions featuring “some-to-all,” “most-to-all,” or subgroup-to-group shifts
  • Weaken questions that attack representativeness (“attendees aren’t typical”)
  • Must Be True questions requiring careful tracking of “some/most/all”
• Common mistakes:
  • Treating “many” as if it means “most”
  • Missing that “not all” is compatible with “some” (it doesn’t mean “none”)
  • Ignoring scope words like “in this study,” “in this city,” “among respondents”

Causal Reasoning: How the LSAT Tests “X Causes Y”

What causal claims assert

A causal claim says that one thing brings about another. Causation is stronger than correlation: it implies that changing the cause would change the effect (all else equal).

The LSAT loves causal reasoning because it’s intuitive in everyday life but logically tricky. Many arguments jump from “X and Y occur together” to “X causes Y,” and that jump is often unjustified.

The big three alternative explanations

When you see “X is associated with Y, so X causes Y,” your mind should automatically consider three classic alternatives:

1. Reverse causation: Y causes X.
2. Common cause: Z causes both X and Y.
3. Coincidence / confounding / selection: the observed relationship is not a direct causal link.

These alternatives power many Weaken answers.

Causal strengthening tools

To strengthen a causal claim, you typically want to:

• eliminate alternative causes (control variables)
• show temporal order (cause happens before effect)
• show a mechanism (how X produces Y)
• show that removing X reduces Y (intervention)

You don’t need scientific sophistication; just recognize the pattern.

Causal weakening tools

To weaken a causal claim, you often:

• provide an alternative cause
• show the effect occurs without the alleged cause
• show the cause occurs without the effect
• point out the data is biased or unrepresentative

Worked example: strengthen vs. weaken

Stimulus: “After the city installed more streetlights, crime decreased. Therefore, the streetlights caused the decrease.”

• A strengthener might say: “Neighboring cities without new streetlights saw no comparable decrease during the same period.” (reduces alternative explanations)
• A weakener might say: “During the same period, the city hired 200 additional police officers.” (alternative cause)

Notice how both answers target the gap: other things could explain the drop.

Exam Focus

• Typical question patterns:
  • Strengthen/weaken arguments that infer causation from correlation
  • Flaw questions: “confuses correlation with causation,” “overlooks alternative causes”
  • Resolve questions where a causal relationship seems inconsistent with data
• Common mistakes:
  • Treating any time-order statement as proof of causation (sequence alone doesn’t prove cause)
  • Picking answers that are merely consistent with the causal claim rather than supportive (strengthen needs impact)
  • Missing that a weakener can attack the evidence (bad data) or the logic (alternative explanation)

Assumptions: The Hidden Bridge Between Premises and Conclusion

What an assumption is

An assumption is a statement the argument needs to be true for the reasoning to work. Arguments rarely state every needed link; they rely on shared beliefs, definitions, or unstated constraints.

Assumptions matter because many LR question types are essentially about finding or manipulating that hidden bridge:

• Necessary Assumption: what must be true
• Sufficient Assumption: what, if added, guarantees the conclusion
• Strengthen/Weaken: often operate by supporting or attacking assumptions
• Flaw: often describes a missing or dubious assumption

Two kinds: necessary vs. sufficient

This distinction is central.

• A necessary assumption is required. If it’s false, the argument collapses.
• A sufficient assumption is a powerful guarantee. If it’s true, the conclusion follows (even if the original argument was weak).

A good way to feel the difference:

• Necessary = “without this, no deal.”
• Sufficient = “with this, done.”

The Negation Test (for necessary assumptions)

For a candidate necessary assumption, you can apply the Negation Test:

1. Negate the answer choice (make it as close to a logical opposite as possible, not a dramatic extreme).
2. Ask: does the argument fall apart?

If negating it destroys the reasoning, it was necessary.

Be careful: negation is not always “add ‘not’.” For example:

• Negation of “some” is “none.”
• Negation of “all” is “not all.”
• Negation of “should” is often “not required/need not,” not “should not.”

Finding assumptions by spotting gaps

Most arguments have a predictable gap pattern:

• Premises talk about one thing; conclusion talks about a slightly different thing (scope shift).
• Premises show a correlation; conclusion asserts causation.
• Premises establish a condition; conclusion asserts a stronger condition.
• Premises offer evidence of one kind (surveys, experts); conclusion makes a broad claim.

Your job is to articulate what must be true for that leap to be valid.

Worked example: necessary vs. sufficient

Stimulus: “All employees who handle cash must pass a background check. Dana handles cash. Therefore, Dana must have passed a background check.”

This argument is actually valid if the premise means what it says and the policy is followed—but notice the subtle assumption: that the rule is enforced.

• A necessary assumption might be: “Dana is an employee (not, say, a customer or volunteer).” If Dana isn’t an employee, the first premise doesn’t apply.
• A sufficient assumption could be: “Anyone who handles cash is an employee.” That would make the conclusion follow even more directly.

Exam Focus

• Typical question patterns:
  • “Which of the following is an assumption on which the argument depends?” (Necessary)
  • “Which of the following, if assumed, allows the conclusion to be properly drawn?” (Sufficient)
  • Strengthen questions where the correct answer looks like it patches a missing link
• Common mistakes:
  • Treating necessary and sufficient as interchangeable (they are not)
  • Negating an answer incorrectly (especially with quantifiers and modal verbs)
  • Choosing an answer that would help the argument but isn’t required (common trap in necessary assumption)

Strengthen and Weaken: Changing the Probability of the Conclusion

The mindset: you’re not proving, you’re shifting likelihood

Strengthen/Weaken questions rarely ask you to make an argument perfectly valid. Instead, you’re asked to make the conclusion more or less supported by the premises.

Think in terms of plausibility: the right answer should noticeably affect how confident you are.

Step-by-step approach

1. Identify the conclusion.
2. Identify the support (premises).
3. Describe the gap—what would need to be true for the premises to justify that conclusion?
4. Evaluate each answer by impact on that gap.

This “gap-first” approach prevents you from being seduced by answers that are on-topic but irrelevant.

Common strengthening moves

A Strengthen answer often:

• supports a key assumption
• eliminates an alternative explanation
• adds a missing premise
• shows the evidence is reliable/representative
• links terms (resolves a definition mismatch)

Common weakening moves

A Weaken answer often:

• attacks a key assumption
• offers a counterexample or alternative cause
• shows the evidence is biased, incomplete, or unrepresentative
• shows the conclusion doesn’t follow even if premises are true

Example: definition mismatch

Stimulus: “This restaurant is the best in town because it has the highest customer satisfaction score.”

Hidden assumption: satisfaction score captures “best.”

• Strengthen: “The satisfaction score is based on a large, randomized sample of diners and correlates strongly with independent expert ratings.”
• Weaken: “The satisfaction score is based only on online reviews from a small group of regulars who received discounts.”

Exam Focus

• Typical question patterns:
  • “Which of the following, if true, most strengthens/weakens the argument?”
  • “The argument would be most strengthened if which of the following were true?” (same task)
  • Weaken questions featuring an alternative explanation as the correct answer
• Common mistakes:
  • Choosing an answer that attacks the conclusion directly but doesn’t touch the reasoning (you must affect support)
  • Confusing “relevant detail” with “logical impact”
  • For weaken: picking something that is merely inconsistent with a premise rather than undermining the inference (unless the inconsistency matters)

Must Be True and Most Strongly Supported: Drawing Careful Inferences

What these questions test

These questions flip the usual direction: you’re not evaluating an argument; you’re extracting what follows from the given statements.

• Must Be True (MBT): the correct answer is logically entailed by the stimulus.
• Most Strongly Supported (MSS): the correct answer has the strongest textual support—even if not 100% guaranteed.

In practice, both reward conservative reading and punish “reasonable speculation.”

How valid inferences are created

Inferences come from:

• combining statements (“A implies B” plus “B implies C”)
• contraposition (in conditional logic)
• careful quantifier reasoning (what “most” allows and forbids)
• eliminating impossibilities

A strong habit is to ask: “What must be true if every statement above is true?”

Trap patterns

Wrong answers often:

• add a new idea not in the stimulus
• strengthen the stimulus’s conclusion (if it’s an argument) but are not entailed
• use extreme language (“never,” “all”) where the stimulus was moderate

Worked example: conservative inference

Stimulus: “All of the files in folder X are backed up. Some files in folder X are encrypted.”

What must be true?

• At least one encrypted file is backed up.

Because “some files in folder X are encrypted” guarantees at least one file is both in X and encrypted, and all files in X are backed up.

Exam Focus

• Typical question patterns:
  • “Which of the following is most strongly supported by the statements above?”
  • “If the statements above are true, which of the following must also be true?”
  • Inference questions built on combining conditionals and quantifiers
• Common mistakes:
  • Bringing in outside knowledge (“in real life, that would mean…”) instead of sticking to what’s stated
  • Mistaking what is possible for what is required
  • Overlooking that “some” guarantees existence (at least one)

Flaw in the Reasoning: Recognizing Common Logical Errors

What flaw questions ask

A Flaw question asks you to describe what is wrong with the reasoning. You are not asked whether the conclusion is true in real life; you’re asked whether the premises logically support it.

The correct answer typically names a recognizable reasoning mistake—often a gap between what’s proven and what’s claimed.

High-frequency flaw categories (conceptually)

Rather than memorizing a long list of labels, focus on the underlying patterns:

1) Sampling and representation flaws

Arguments based on surveys or studies often assume the sample represents the population.

• Flaw: drawing a conclusion about all people from a biased subset.

2) Causal flaws

As covered earlier:

• confusing correlation with causation
• ignoring alternative causes
• mixing up cause and effect

3) Conditional fallacies

• affirming the consequent
• denying the antecedent

4) Equivocation (shifting meaning)

An argument uses a key term in two different senses.

Example: “A ‘theory’ is just a guess. Evolution is a theory. Therefore evolution is just a guess.” (Different meanings of “theory.”)

5) Part-to-whole / whole-to-part

• what’s true of a part assumed true of the whole
• what’s true of the whole assumed true of each part

6) Normative/conceptual leaps

Moving from a descriptive claim (“people do X”) to a prescriptive conclusion (“people should do X”) without a value premise.

Worked example: identifying the flaw precisely

Stimulus: “Whenever the school increases security, student anxiety rises. So the security increase causes anxiety.”

Flaw: assumes causation from correlation; may ignore that anxiety could cause security increases (reverse causation) or a third factor (recent incidents) could cause both.

A good flaw description will mention what the author failed to consider.

Exam Focus

• Typical question patterns:
  • “The reasoning is flawed because the argument…”
  • “Which of the following most accurately describes an error in the reasoning?”
  • Flaw questions paired with classic patterns: causation, sampling, equivocation
• Common mistakes:
  • Choosing an answer that criticizes the topic rather than the logic
  • Picking a true criticism that doesn’t match the argument’s actual move
  • Missing subtle shifts in meaning (same word, different sense)

Method of Reasoning and Role/Function: Describing What the Argument Does

Method of reasoning: the argument’s “moves”

A Method of Reasoning question asks for a description of how the argument proceeds—what it does with evidence to reach a conclusion.

Common methods include:

• drawing a general conclusion from examples
• using elimination (ruling out alternatives)
• applying a general rule to a specific case
• using an analogy
• explaining a phenomenon with a hypothesis

The key is to summarize structure without getting lost in topic details.

Role/Function: what a specific sentence is doing

A Role or Function question asks what a particular statement does in the argument:

• main conclusion
• premise
• intermediate conclusion
• concession
• counterpoint/opposing view
• background information

A powerful habit is to reread the indicated sentence and ask: “If I removed this sentence, what would change?” If removing it destroys support, it’s likely a premise. If removing it removes what’s being argued for, it’s likely the conclusion.

Worked example: function recognition

Stimulus: “Some critics say the policy will increase costs. However, the cost projections ignore savings from reduced turnover. Therefore the policy will not increase overall costs.”

• “Some critics say …” = opposing view/background
• “However … ignore savings …” = premise undermining the critics
• “Therefore …” = conclusion

Exam Focus

• Typical question patterns:
  • “The argument proceeds by…” (method)
  • “The statement that X plays which one of the following roles?”
  • “The author mentions X in order to…”
• Common mistakes:
  • Describing content instead of structure (“talks about economics”) rather than the reasoning move
  • Confusing a premise that attacks an opposing view with the author’s own conclusion
  • Missing the main conclusion when the argument includes multiple viewpoints

Parallel Reasoning and Parallel Flaw: Matching Structure, Not Topic

What parallel questions test

Parallel Reasoning asks you to find an answer choice whose reasoning has the same logical form as the stimulus. Parallel Flaw asks for the same logical mistake.

These are pattern-recognition questions: the topic will change, but the structure must match.

How to abstract an argument

To match structure, translate the stimulus into a simple template using placeholders:

• Identify conclusion and premises.
• Identify any conditional logic and quantifiers.
• Note whether the reasoning is valid or flawed.

For example, this flawed form:

• If A then B.
• B.
• Therefore A.

is a classic template you can match quickly.

What “match” really means

A correct parallel answer must match:

• the number of premises and their roles (support vs. intermediate)
• the logical operators (conditional, causal, quantifiers)
• the direction of inference (including invalid direction)

It does not need to match:

• subject matter
• tone
• plausibility

Worked example: parallel flaw

Stimulus: “If the engine overheats, the warning light turns on. The warning light is on, so the engine must be overheating.”

This is affirming the consequent.

Correct parallel flaw will also take the form:

• If A then B. B. Therefore A.

Even if the answer is about pets, taxes, or weather, that logical skeleton must be the same.

Exam Focus

• Typical question patterns:
  • “Which of the following arguments is most similar in its reasoning?”
  • “Which of the following exhibits flawed reasoning most similar to that above?”
  • Parallel questions built around conditional fallacies and causal gaps
• Common mistakes:
  • Matching on topic similarity instead of logical structure
  • Ignoring whether the stimulus is valid or flawed (parallel flaw requires flawed)
  • Overlooking quantifiers or negations that change the form

Principle Questions: Using General Rules as Bridges

Two major kinds of principle tasks

Principle questions come in two common forms:

1. Principle–Justify (Principle that supports the argument)
   • You’re given an argument and asked for a general principle that, if true, would strengthen/justify it.
2. Principle–Conform (Principle that the argument follows)
   • You’re given an argument and asked which general principle the argument exemplifies.

They feel similar but behave differently.

Principle–Justify: a sufficient-assumption cousin

When a principle is supposed to support the argument, it often functions like a sufficient assumption—a broad rule that connects the premises to the conclusion.

You want a principle that:

• is broad enough to cover the case in the stimulus
• directly links the evidence type to the conclusion type
• does not introduce irrelevant conditions

Principle–Conform: describing what rule is being used

When asked which principle the argument follows, you are summarizing the rule the argument is implicitly applying.

A good method:

• Paraphrase the argument as “Whenever [premise pattern], [conclusion pattern].”
• Then find an answer that states that pattern as a general rule.

Worked example: principle–justify

Stimulus: “The committee should reject the proposal because it would violate the organization’s privacy policy.”

A supporting principle might be: “Any proposal that violates the organization’s privacy policy should be rejected by the committee.”

This bridges: policy violation → should reject.

Exam Focus

• Typical question patterns:
  • “Which of the following principles, if valid, most helps to justify the reasoning?”
  • “The reasoning conforms to which of the following general principles?”
  • Principles that connect descriptive facts to normative conclusions
• Common mistakes:
  • Picking a principle that sounds morally appealing but doesn’t logically connect premises to conclusion
  • Choosing a principle that is too strong (“always”) or too weak (“sometimes”) for the argument’s force
  • Missing necessary details (e.g., the principle talks about “illegal” while the stimulus is about “unethical”)

Explain/Resolve the Paradox: Making Two Facts Fit Together

What these questions look like

A Resolve/Explain question gives you a situation where facts seem inconsistent:

• a surprising result
• a conflict between two findings
• an apparent contradiction

Your job is to find an answer that, if true, makes the facts compatible.

The key mindset: you are not strengthening an argument

These questions often don’t have a conclusion at all. Treat the stimulus like a puzzle:

• Fact A suggests one thing.
• Fact B suggests the opposite.

A good resolution adds a missing piece showing why both can be true.

Common resolution strategies

Resolutions often:

• introduce a hidden difference between groups, time periods, or definitions
• show that one measure doesn’t track what you think it tracks
• add a third factor that explains the surprising outcome

Worked example: resolving

Stimulus: “A new medicine performed well in laboratory tests, but in clinical trials it was no more effective than a placebo.”

A resolving answer might say: “The medicine breaks down quickly in the human digestive system, but remained stable in laboratory conditions.”

This explains why lab success doesn’t translate to real-world trials.

Exam Focus

• Typical question patterns:
  • “Which of the following, if true, most helps to resolve the apparent discrepancy?”
  • “Which option best explains the surprising finding?”
  • Stimuli that compare studies, time periods, or measurement methods
• Common mistakes:
  • Choosing an answer that takes sides (supports Fact A or Fact B) instead of reconciling them
  • Treating it like a weaken question and making the discrepancy worse
  • Missing that a resolution can be very specific (small detail that explains a big mismatch)

Evaluate the Argument: Testing What Would Matter

What “evaluate” really means

An Evaluate question asks: which piece of information would be most useful to determine whether the argument’s conclusion is valid?

This is different from strengthen/weaken because you’re not asked to push the argument in one direction—you’re asked what would help you decide.

The “yes/no” diagnostic

A classic property of correct Evaluate answers: if the answer to the question were yes, it would affect the argument one way; if no, it would affect it the other way.

So the right answer often looks like:

• “Whether …”
• “Which of the following would be most relevant to determining …”

How to find the right evaluation point

1. Identify the argument’s core gap or assumption.
2. Ask what fact would confirm vs. undermine that assumption.
3. Look for an answer that directly targets that pivot.

Worked example: evaluate

Stimulus: “This tutoring program improves scores because students who used it scored higher than those who didn’t.”

Key gap: selection bias—maybe stronger students chose tutoring.

Good evaluation question: “Did the students who used the program have similar prior scores to those who did not?”

• If yes (similar), program looks more causal.
• If no (users were already stronger), the conclusion is weaker.

Exam Focus

• Typical question patterns:
  • “Which of the following would be most useful to know in order to evaluate the argument?”
  • “The answer to which question would help determine whether…?”
  • Arguments relying on causal inferences or representativeness
• Common mistakes:
  • Picking an answer that would be interesting but doesn’t change your judgment
  • Choosing something that strengthens only (not a two-way diagnostic)
  • Evaluating a minor detail rather than the argument’s central leap

Numbers, Percentages, and Comparisons: Avoiding Statistical Traps

Why “mathy” stimuli are usually logical traps

LR doesn’t require advanced computation, but it frequently uses numerical language to test reasoning:

• percentages vs. totals
• averages
• “increased by” vs. “increased to”
• comparisons across different baselines

The common thread: you’re being tested on whether you notice what’s actually being compared.

Percent vs. absolute number

A percentage can rise even if the total falls, and vice versa.

Example idea:

• If a company downsizes, the percentage of employees in management could increase even if the number of managers decreases.

So if an argument treats a percent change as if it proves an absolute change (or the reverse), that’s a flaw.

Averages can hide distribution

An argument may conclude that “typical” improved because the average improved, but the improvement could come from a few extreme cases.

This often connects to representativeness: average is one metric, not the whole story.

“More than” comparisons and shifting denominators

Watch for comparisons like:

• “more accidents this year” without controlling for “more drivers this year”
• “higher rate” vs. “higher count”

Rates require denominators.

Worked example: denominator shift

Stimulus: “This year there were more bicycle accidents than last year, so cycling has become more dangerous.”

Missing piece: number of cyclists or miles ridden. If cycling grew dramatically, the accident rate might have fallen.

A strong weakener: “The number of cyclists doubled this year compared to last year.”

Exam Focus

• Typical question patterns:
  • Weaken/Flaw questions built on percent vs. total confusions
  • Evaluate questions asking about base rates/denominators
  • MSS questions where the correct inference is modest (“the rate may not have increased”)
• Common mistakes:
  • Treating “higher number” as “higher probability” automatically
  • Ignoring what population the percentage refers to
  • Over-interpreting an average as if it describes most individuals

Common Answer Choice Traps (and How to Think Past Them)

Why traps work

Wrong answers aren’t random; they’re designed to reward shallow processing:

• they repeat keywords from the stimulus
• they sound generally reasonable
• they address the topic but not the logic

The LSAT is testing whether you can keep your attention on the task (the stem) and the structure (premise-to-conclusion).

Frequent trap styles

1) Out of scope

The answer introduces a new issue not discussed. It may be relevant in real life, but it doesn’t connect to the argument.

2) Too strong / too weak

• Too strong: uses “all/never” when the stimulus doesn’t support that force.
• Too weak: uses “might/could” in a way that doesn’t affect a strong claim.

3) Restating a premise

In strengthen/weaken, an answer that just repeats a premise doesn’t add support. It feels comfortable because it’s familiar, but it doesn’t move the argument.

4) Attacking a side point

Some arguments contain multiple claims. Trap answers weaken or strengthen a minor claim while leaving the main conclusion’s support unchanged.

Worked example: premise restatement

Stimulus: “The park should be renovated because it is unsafe.”

Trap strengthener: “The park is unsafe.” (premise restated)

Real strengthener would add support for why renovation solves unsafety or why unsafety warrants renovation.

Exam Focus

• Typical question patterns:
  • Strengthen/weaken questions where 2–3 answers are “on-topic” but non-impactful
  • MBT questions where traps use extreme language
  • Flaw questions where traps describe a different flaw than the one present
• Common mistakes:
  • Selecting answers that reuse stimulus language without checking logical function
  • Overvaluing “common sense” plausibility instead of relevance to the reasoning
  • Failing to re-check the conclusion when multiple viewpoints appear

Putting It Together: A Repeatable Process for Any LR Question

A disciplined workflow

Across question types, you can apply a consistent sequence:

1. Read for structure: identify conclusion, premises, and any intermediate conclusions.
2. Name the gap (if it’s an argument): what must be true, what’s assumed, what alternative explanations exist.
3. Label the task: infer, fix, attack, describe, reconcile, match.
4. Predict what would work before going to choices (a “prephrase”).
5. Use elimination with a reason: wrong because out of scope, wrong force, wrong direction, wrong role.

This matters because LR rewards consistency more than flashes of insight. Your goal is to reduce each question to a familiar type of move.

Mini worked set: same stimulus, different tasks

Stimulus: “Residents near the airport report more sleep problems than residents elsewhere in the city. Therefore, airplane noise causes sleep problems.”

• Weaken: “Residents near the airport also tend to work night shifts.” (alternative cause)
• Strengthen: “After a new nighttime flight restriction, sleep problems in the area decreased.” (intervention)
• Necessary assumption: “The surveys accurately measured sleep problems in both groups.” (measurement validity)
• Flaw: “Confuses correlation with causation; fails to rule out alternative explanations.”

Same facts, different tasks—your stem determines your move.

Exam Focus

• Typical question patterns:
  • Mixed sets where similar stimuli appear across different tasks
  • “Which of the following, if assumed…” vs. “Which of the following, if true…”
  • Questions that hinge on identifying the main conclusion quickly
• Common mistakes:
  • Skipping the prephrase and letting answer choices dictate your thinking
  • Focusing on content details at the expense of argument structure
  • Treating “if true” as “must be true” (different standards)