LSAT Logical Reasoning: Building, Analyzing, and Evaluating Arguments

What an LSAT “Argument” Is (and How to Find Its Parts)

Logical Reasoning (LR) is built around short passages that contain arguments—attempts to prove something. Your core job is to understand what the author is trying to establish and how they’re trying to establish it. Almost every LR question becomes easier once you can reliably separate conclusion from premises and notice any hidden assumptions.

An argument (in LSAT terms) is a set of statements where at least one statement (the conclusion) is claimed to follow from other statements (the premises). The LSAT also includes passages that are not arguments—many are explanations, descriptions, or sets of facts. Treating a non-argument like an argument is a common source of confusion.

Conclusion vs. premises

• A conclusion is the main claim the author is trying to get you to accept.
• A premise is a reason offered in support of the conclusion.

A practical way to think about it: if you asked the author “Why should I believe that?”, the answer is (usually) the premises. If you asked “So what?”, the answer is (usually) the conclusion.

Indicator words can help, but they’re not foolproof.

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

The LSAT frequently tests whether you can find the conclusion even when there are no indicators—or when indicator-like words are used in nonstandard ways.

Main conclusion vs. intermediate conclusion

Many arguments have layers. An intermediate conclusion is a claim that is supported by some premises and then used as a premise to support the main conclusion.

When you see a chain, your job is to identify the top of the chain (the main conclusion). This matters because many question types (Strengthen, Weaken, Assumption) care about the gap between premises and the main conclusion—not some earlier sub-conclusion.

Background facts, context, and “scope”

Not every sentence is doing logical work. LR passages often include:

• Background: scene-setting information that makes the argument understandable.
• Concessions: “Although X, nevertheless Y.” (X is often not the author’s final view.)
• Counterarguments: someone else’s view, presented to be rejected.

Recognizing what the author endorses versus what the author merely reports is essential—especially in Method, Flaw, and Role questions.

Argument vs. explanation

An explanation aims to account for why something happened, not to prove that it happened. For example:

• “The streets are wet because it rained overnight.”

This can be an argument (“It rained, therefore the streets are wet”) or an explanation (“Given that the streets are wet, the cause is rain”). The direction of support is what determines which it is.

Worked example: identifying structure

Passage: “City traffic has increased since the new shopping center opened. Therefore, the shopping center caused the increase in traffic.”

• Premise: Traffic increased since the center opened.
• Conclusion: The shopping center caused the increase.

Notice the conclusion goes beyond the premise: “since” (time correlation) is used to infer “caused.” That gap is exactly what later questions will exploit.

Exam Focus

• Typical question patterns:
  • “Which of the following is the conclusion?” or “The author’s reasoning proceeds by…”
  • “The statement ‘X’ plays which role?” (premise, conclusion, objection, etc.)
  • Passages that include competing viewpoints and ask what the author would agree/disagree with
• Common mistakes:
  • Treating a reported viewpoint (someone else’s claim) as the author’s conclusion
  • Confusing a causal explanation with an argument (mixing up what is being supported)
  • Stopping at an intermediate conclusion and missing the main conclusion

Reading for Reasoning: A Repeatable Process for LR Passages

You don’t need to read LR passages like literature; you need to read them like a critic of reasoning. The goal is to build a quick internal model: what’s the conclusion, what are the premises, and what assumption must be true for the premises to actually support the conclusion.

A practical 4-step approach

1. Find the conclusion: Ask, “What is the author trying to prove?”
2. List the explicit premises: Identify what facts or claims are offered as support.
3. Describe the gap: In plain language, state what would have to be true for the premises to make the conclusion likely.
4. Predict what would help or hurt: What would strengthen the link? What would undermine it?

This “gap-focused” reading matters because many LR questions are variations on the same task: modify or evaluate the argument by targeting that gap.

Why “prephrasing” helps (and how to do it without overcommitting)

Prephrasing means making a tentative prediction of what a correct answer might say before you look at the options. This combats a major LR trap: attractive answer choices that are true-sounding but irrelevant.

A good prephrase is usually abstract and flexible:

• “The argument assumes there isn’t some other cause.”
• “We need a connection between the survey sample and the whole population.”

A bad prephrase is overly specific:

• “The survey was conducted on Tuesday at 3 pm.”

Matching the question stem to the task

Question stems tell you what role the correct answer must play. For example:

• “Which of the following, if true, most strengthens…?” asks you to add support.
• “The argument depends on which assumption?” asks you to find a required bridge.
• “Which of the following is most strongly supported?” asks you to infer from the passage without adding new information.

A powerful habit is to translate stems into a short command:

• Strengthen: “Help the conclusion follow.”
• Weaken: “Attack the link.”
• Necessary assumption: “Must be true for the argument to work.”
• Sufficient assumption: “If true, guarantees the conclusion.”

Exam Focus

• Typical question patterns:
  • Stems that hinge on subtle words: “most strongly supported” vs. “must be true”; “helps” vs. “requires”
  • “EXCEPT” stems (find the one that does not do the asked job)
  • Questions where the passage is short but the answer choices are long and tempting
• Common mistakes:
  • Reading for topic rather than reasoning (remember: LR tests logic, not facts)
  • Ignoring stem keywords (especially “depends on,” “required,” “if true”)
  • Falling for answer choices that are true in real life but not relevant to the argument

Conditional Reasoning: If–Then Logic and the Contrapositive

A major portion of LR involves statements that set rules or relationships: conditional statements. These are “if–then” claims that say one condition is sufficient to guarantee another.

The basics: sufficient vs. necessary

In a conditional statement “If A, then B”:

• A is sufficient for B. If A happens, that’s enough to ensure B.
• B is necessary for A. A cannot happen without B.

So:

• “If you are a cat, then you are a mammal.”
  • Being a cat is sufficient for being a mammal.
  • Being a mammal is necessary for being a cat.

Students often mix these up because everyday language uses “only if,” “unless,” and “requires” in ways that feel backward. LR rewards you for being precise.

The contrapositive (the one valid inference)

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

• “If not B, then not A.”

This is crucial because many questions test whether you can spot invalid “mirror-image” moves.

Valid:

• If A → B
• Not B → Not A

Invalid:

• Affirming the consequent: B → A (wrong)
• Denying the antecedent: Not A → Not B (wrong)

Translating common wording

Here are high-frequency translations:

• “A only if B” means A → B (B is necessary)
• “A if B” means B → A
• “A unless B” usually means: If not B → A (equivalently, If not A → B)
• “A requires B” means A → B
• “A is sufficient for B” means A → B
• “A is necessary for B” means B → A

Conditional chains and combining rules

LR often gives multiple conditionals you can link:

• If A → B
• If B → C
  Then:
• If A → C

Be careful not to chain across a break. If you have “A → B” and “C → D,” you can’t conclude anything connecting A to D.

Worked example: spotting a fallacy

Claim: “If someone is a great lawyer, they can reason well. Dana can reason well, so Dana is a great lawyer.”

• Conditional: Great lawyer → Reasons well
• Argument uses: Reasons well → Great lawyer (affirming the consequent)
  This is invalid: many people reason well without being great lawyers.

Exam Focus

• Typical question patterns:
  • Flaw questions that hinge on confusing sufficient and necessary conditions
  • Must Be True / Inference questions requiring a contrapositive
  • Parallel reasoning questions where you match conditional structure, not topic
• Common mistakes:
  • Treating “only if” as if it introduces the sufficient condition (it introduces the necessary condition)
  • Taking the converse as equivalent to the original conditional
  • Forgetting that “unless” typically creates a “not” on one side of the rule

Quantifiers and Formal Logic in Natural Language (All, Most, Some)

Not all logical relationships are “if–then.” Many arguments use quantifiers like “all,” “most,” and “some.” These words have precise logical implications that the LSAT expects you to respect.

Universal vs. existential claims

• Universal: “All,” “every,” “none,” “no.”
  • “All A are B” rules out any A that is not B.
• Existential: “Some,” “at least one,” “there exists.”
  • “Some A are B” only guarantees one or more; it does not imply many.

A key asymmetry: universal statements are strong and easy to disprove with one counterexample; existential statements are weak and hard to disprove.

“Most” and what it does (and does not) mean

Most means “more than half.” It does not mean “almost all.” From “Most A are B,” you cannot conclude:

• “All A are B” (too strong)
• “Most B are A” (reverses the relationship)
• “Some A are not B” (actually, you can conclude this is possible, but not required; if most are B, some could still be all B if A is small? Careful: if “most A are B,” then at least one A is B, and it’s compatible with some A not B, but not logically forced unless A has at least two members; LSAT typically treats groups as large enough, but you should still avoid over-inference.)

A safe inference from “Most A are B” is:

• “Some A are B.”

Negations: what it takes to contradict a claim

Understanding negation helps with Must Be True and also with Assumption work.

• Negation of “All A are B” is “Some A are not B.”
• Negation of “Some A are B” is “No A are B.”
• Negation of “Most A are B” is “At most half of A are B.”

You don’t need heavy formalism, but you do need to know what would actually contradict the statement.

Quantifiers and argument strength

LR often tests arguments that slide from weak evidence to strong conclusions:

• Premise: “Some nutrition studies show…”
• Conclusion: “Therefore, this diet is the best for everyone.”

The mismatch in strength is usually the vulnerability.

Worked example: quantifier mismatch

Passage: “Some members of the committee opposed the policy. Therefore, the committee opposed the policy.”

• Premise: Some opposed.
• Conclusion: The whole committee opposed.
  This mistakenly treats “some” as “all/most.”

Exam Focus

• Typical question patterns:
  • Flaws that hinge on shifting from “some” to “all,” or from “most” to “all”
  • Inference questions that ask what must be true given “all/none” statements
  • Strengthen/weaken questions where the key is matching the quantifier strength
• Common mistakes:
  • Treating “some” as if it implies “many”
  • Reversing quantified relationships (“Most A are B” does not imply “Most B are A”)
  • Negating incorrectly (especially “most”)

Causal Reasoning: Correlation, Alternative Causes, and Mechanisms

Causal arguments are everywhere on the LSAT because they’re extremely common in real life—and extremely easy to do badly.

A causal claim says one thing produces or influences another (X causes Y). The LSAT tests whether the evidence actually supports that directional, explanatory link.

The classic correlation-to-causation problem

A very common structure is:

• Premise: X and Y occur together (or change together).
• Conclusion: X caused Y.

But correlation can arise for many reasons:

1. Reverse causation: Y causes X.
2. Common cause: Z causes both X and Y.
3. Coincidence: no causal connection.

LR questions often ask you to strengthen or weaken by addressing these alternatives.

Causal strengthening: what really helps

To strengthen “X causes Y,” you can:

• Rule out alternative causes (show Z isn’t responsible).
• Show temporal order (X happens before Y).
• Show a mechanism (how X could produce Y).
• Use comparison or control: when X is present, Y increases; when X is absent, Y decreases.

Causal weakening: where to attack

To weaken “X causes Y,” you can:

• Provide an alternative cause.
• Show Y occurs without X.
• Show X occurs without Y.
• Show the timing doesn’t fit.

“Causal explanation” vs “causal conclusion”

Sometimes the conclusion isn’t “X caused Y,” but rather “X is the explanation for Y.” The same vulnerabilities apply, but the wording can trick you into thinking it’s merely describing.

Worked example: weakening a causal argument

Passage: “After the city installed more streetlights, crime decreased. Therefore, the additional streetlights reduced crime.”
Ways to weaken:

• Show another anti-crime initiative began at the same time (alternative cause).
• Show crime decreased in similar cities without new streetlights (common trend).
• Show crime reports decreased because reporting changed, not because crime changed (measurement issue).

Notice what does not weaken well: “Streetlights are expensive.” That may be true, but it doesn’t target whether they reduced crime.

Exam Focus

• Typical question patterns:
  • Strengthen/weaken arguments based on before/after comparisons
  • Resolve the paradox questions where two facts seem inconsistent due to causal assumptions
  • Assumption questions where the missing link is “no other cause” or “same measurement method”
• Common mistakes:
  • Attacking the conclusion with an irrelevant complaint (cost, morality) instead of causality
  • Forgetting reverse causation as a major alternative
  • Ignoring selection effects (who chose to do X may already differ)

Assumptions: Necessary vs. Sufficient (and How to Test Them)

Assumption questions are among the most important in LR because they train you to see the invisible logic holding an argument together.

An assumption is an unstated claim that the author relies on. If it’s false, the argument collapses or at least loses support.

Necessary assumptions (required assumptions)

A necessary assumption is something the argument must rely on. If it fails, the argument cannot succeed as stated.

These stems often say:

• “The argument depends on which of the following?”
• “Which assumption is required?”
• “Which must be assumed?”

The negation test (why it works)

A reliable method for necessary assumptions is the negation test:

1. Take an answer choice.
2. Negate it (carefully—not always by adding “not,” but by making it meaningfully opposite).
3. Ask: if the negation were true, would the argument fall apart?

• If yes, it was necessary.
• If no, it wasn’t necessary.

Example of careful negation:

• Original: “Some of the cars are electric.”
• Negation: “None of the cars are electric.”

Sufficient assumptions (justifying the conclusion)

A sufficient assumption is strong enough that, if you add it to the premises, the conclusion follows.

These stems often say:

• “Which of the following, if assumed, allows the conclusion to be properly drawn?”
• “Which principle, if valid, justifies the reasoning?”

Sufficient assumptions tend to be more powerful and more “sweeping” than necessary assumptions because they must guarantee the conclusion.

Bridging the gap: the central skill

Most assumption tasks reduce to:

• Premises talk about A.
• Conclusion talks about B.
• Missing link is a rule connecting A to B.

A common pattern is the scope shift:

• Premise: about a sample.
• Conclusion: about a population.
  Necessary assumption: the sample represents the population.

Worked example: necessary vs sufficient

Passage: “All trained service dogs are calm in crowds. Riley is calm in crowds. Therefore, Riley is a trained service dog.”

• This is invalid (affirming the consequent).
• A sufficient assumption that would make it valid: “Only trained service dogs are calm in crowds.” (That gives Calm → Trained.)
• A necessary assumption might be something like: “Riley is a dog.” Even if you add “Riley is a dog,” the argument is still invalid—but the author is at least assuming Riley is in the relevant category.

This shows the difference: necessary assumptions keep the argument from being nonsensical; sufficient assumptions fix the logic.

Exam Focus

• Typical question patterns:
  • “Depends on” stems (necessary) vs. “allows/justifies” stems (sufficient)
  • Arguments with a clear gap between a property and a classification (common in sufficient assumption)
  • Assumptions about representativeness, causality, or definitions
• Common mistakes:
  • Using the negation test on sufficient assumption questions (it’s designed for necessary assumptions)
  • Picking an answer that strengthens but doesn’t have to be true (confusing Strengthen with Necessary Assumption)
  • Negating sloppily (turning “some” into “not all” rather than “none,” etc.)

Strengthen and Weaken: Targeting the Argument’s Vulnerability

Strengthen and Weaken questions are the clearest place where “gap thinking” pays off. You’re not being asked to find a true statement; you’re being asked to find the statement that most affects the argument’s support.

A strengthener increases the likelihood that the conclusion follows from the premises. A weakener decreases it.

The “most” standard: you’re choosing the best of five

These questions are comparative. Several answers may be somewhat relevant; your job is to pick the one that has the biggest impact given the argument’s logic.

Common strengthen moves

• Support a missing assumption (especially about representativeness, causality, or definitions).
• Provide evidence that the stated premise pattern is reliable.
• Rule out an alternative explanation.

Common weaken moves

• Provide counterevidence to the conclusion.
• Show the premise is unreliable (bad data, biased sample, ambiguous term).
• Introduce an alternative explanation that better accounts for the premises.

Beware of “out of scope” truth

A frequent trap answer is something that is plausible or factually true but doesn’t connect to the argument.

A useful check: after reading an answer choice, ask “Does this change how well the premises support the conclusion?” If it doesn’t, it’s not the right answer—even if it’s interesting.

Worked example: strengthening with representativeness

Passage: “In a survey of 200 residents of Town X, 70% said they support the new park. Therefore, most residents of Town X support the new park.”
A strong strengthener: “The survey respondents were selected randomly from all adult residents of Town X.”
Why it works: it reduces sampling bias and connects the sample to the population.

A weak or irrelevant strengthener: “The new park would include a playground.” That may make support more likely in the world, but it doesn’t address whether the survey supports the conclusion.

Exam Focus

• Typical question patterns:
  • Strengthen/weaken causal claims, survey claims, and policy recommendations
  • Weaken questions where the best answer introduces an alternative cause or shows the effect without the cause
  • Strengthen questions where the best answer rules out a key objection
• Common mistakes:
  • Choosing an answer that is consistent with the conclusion but doesn’t add support
  • Overvaluing extreme language (“proves,” “guarantees”) on Strengthen questions
  • Missing the argument’s real vulnerability and attacking something the author never relied on

Common Logical Flaws (and How the LSAT Dresses Them Up)

Flaw questions reward you for recognizing recurring patterns of bad reasoning. The LSAT rarely invents new mistakes; it repeats classic ones under different topics.

A flaw is a defect in reasoning—an invalid move, an unsupported leap, or a confusion of concepts. The challenge is that flaw descriptions in answer choices are abstract; you must match the abstract description to the concrete argument.

Confusing necessary and sufficient conditions

We covered the formal side earlier; here’s how it appears as a flaw:

• Treating a necessary condition as if it guarantees the outcome.
• Treating a sufficient condition as if it’s required.

Mistaken reversal and mistaken negation

Two named versions of conditional confusion:

• Mistaken reversal: A → B, so B → A.
• Mistaken negation: A → B, so not A → not B.

Equivocation: shifting meaning of a key term

Equivocation happens when an argument uses the same word in two different senses.

Example: “A ‘theory’ is just a guess. Evolution is a theory. Therefore, evolution is just a guess.”
Here “theory” shifts from scientific theory (well-supported framework) to casual “guess.”

Part-to-whole and whole-to-part

• Composition: What is true of parts is assumed true of the whole.
• Division: What is true of the whole is assumed true of the parts.

Example (composition): “Each component is lightweight, so the machine is lightweight.” (Maybe, but not guaranteed.)

Sampling and survey flaws

Common problems:

• Biased sample (not representative)
• Small or self-selected sample
• Leading questions
• Unreliable measurement

Causal flaws

Beyond correlation vs causation, common causal flaws include:

• Assuming a cause is the only cause.
• Confusing absence of evidence with evidence of absence.
• Ignoring that an intervention may have side effects that change behavior (people adapt).

Circular reasoning

Circular reasoning happens when the conclusion is assumed in the premises.

Example: “This law is unjust because it is unfair.” If “unjust” and “unfair” are effectively the same claim, the premise doesn’t provide independent support.

False dilemma (either-or)

A false dilemma limits options to two when more exist.

Example: “Either we ban cars downtown or traffic will get worse.” Perhaps there are other solutions.

Attacking the person (ad hominem) and irrelevant appeals

Sometimes an argument dismisses a claim by attacking the speaker or by appealing to popularity/tradition rather than evidence.

Important nuance: simply mentioning a person isn’t ad hominem. It’s ad hominem when the person’s character is used as a substitute for engaging the argument.

Worked example: matching a flaw description

Passage: “No experienced pilot would ignore a preflight checklist. Kim ignored the preflight checklist. So Kim is not an experienced pilot.”
This is actually valid via contrapositive:

• Experienced → Not ignore
• Ignore → Not experienced

A common test move is to offer an answer choice accusing it of “denying the antecedent” or “mistaken reversal.” You must check the structure before accepting a tempting flaw label.

Exam Focus

• Typical question patterns:
  • “The reasoning is flawed because…” with abstract descriptions
  • Flaw questions built around causality, quantifiers, and conditional logic
  • Parallel Flaw questions (match the same mistake in a new argument)
• Common mistakes:
  • Picking a flaw that sounds bad in general but doesn’t describe what happened in the passage
  • Missing that an argument is valid (and assuming every argument must be flawed)
  • Confusing a premise you disagree with for a reasoning flaw (LR tests validity of support, not whether you like the claim)

Inference Questions: Must Be True vs. Most Strongly Supported

Inference-style questions look similar but have different standards of proof. The LSAT is testing how carefully you can draw conclusions from given information without adding assumptions.

Must Be True (MBT)

A Must Be True question asks for a statement that is forced by the passage—true in every situation where the passage is true.

Good MBT technique:

• Translate/organize the facts.
• Combine conditionals using valid moves (especially contrapositive).
• Avoid answer choices that go beyond the given scope (stronger quantifiers, new groups, new causes).

Most Strongly Supported (MSS)

A Most Strongly Supported question asks for the statement that has the best support, even if it’s not 100% guaranteed.

This is a subtle but important difference. MSS correct answers are often modest, carefully worded, and “safer.” If an answer sounds like it’s making a big leap, it’s usually wrong.

Recognizing “safe” language

For MSS especially, weaker language can be a clue:

• “Some,” “often,” “at least one,” “likely,” “tends to”

But don’t treat this as a rule. Weak language that’s irrelevant is still wrong.

Worked example: MBT with a contrapositive

Facts:

1. “All employees who handle cash must complete training.”
2. “No interns have completed the training.”
   Question: What must be true?

From (1): Handle cash → Completed training.
Contrapositive: Not completed training → Not handle cash.
From (2): Intern → Not completed training.
So: Intern → Not handle cash.
A must-be-true statement: “No interns handle cash.”

Exam Focus

• Typical question patterns:
  • MBT questions built from 2–4 conditional statements
  • MSS questions where several answers are plausible, but only one is tightly supported
  • Inference questions that embed quantifiers (“most,” “some,” “none”)
• Common mistakes:
  • Treating MSS as MBT (demanding certainty when the question asks for best support)
  • Adding real-world assumptions (“interns usually…”) not in the passage
  • Missing a contrapositive and therefore missing the key inference

Method, Role, and Principle Questions: Thinking About the Argument as a Machine

Some LR questions ask you to step back and describe the reasoning rather than evaluate it. These are often easier once you treat the argument like a machine with inputs (premises) and an output (conclusion).

Method of reasoning

Method questions ask what the argument does—often in abstract terms.

Common method descriptions include:

• Drawing a conclusion from evidence
• Offering an explanation and ruling out an alternative
• Applying a general rule to a specific case
• Using an analogy

To answer method questions, you typically:

1. Identify the conclusion.
2. Identify the key support.
3. Summarize the move in one sentence.
4. Match that to an answer choice.

Role of a statement

Role questions ask what a specific statement does in the argument. A sentence can function as:

• A premise
• The main conclusion
• An intermediate conclusion
• A counterpremise (objection)
• Background information

A reliable test: ask “If I remove this sentence, what changes?” If the argument loses support, it was likely a premise. If the argument loses its point, it was likely the conclusion.

Principle questions

A principle is a general rule. Principle questions come in two common forms:

1. Principle (justify): Find a principle that, if true, makes the argument valid.
2. Principle (conform): Find a principle that the argument follows.

In both cases, you want a principle that matches the argument’s structure—not its topic. Many wrong answers are either too strong (would justify more than the argument claims) or mismatched (about a different kind of situation).

Worked example: principle that justifies

Argument: “This restaurant advertised that its soup is vegetarian. It contains chicken stock. Therefore, the restaurant’s advertisement is misleading.”
A justifying principle might be: “If a product is advertised as having a property that it in fact lacks, the advertisement is misleading.”
This connects the premise (advertised vegetarian, actually not) to the conclusion (misleading).

Exam Focus

• Typical question patterns:
  • Method questions with abstract answer choices that sound similar
  • Role questions targeting a single line in a multi-step argument
  • Principle questions that require matching a general rule to a specific scenario
• Common mistakes:
  • Choosing an answer that restates the conclusion instead of describing the reasoning
  • Confusing “conforms to” with “would justify” (following a principle vs needing it)
  • Picking a principle that is morally appealing but logically mismatched

Resolving Paradoxes and Evaluating Explanations

These questions present a situation that seems inconsistent and ask you to resolve it. They’re less about formal validity and more about finding the missing fact that makes both statements possible.

A paradox (or apparent discrepancy) is a pair of facts that don’t seem to fit together. The correct answer provides information that, if true, shows there’s no real contradiction.

How to approach discrepancy questions

1. Identify the two facts that conflict.
2. Ask what assumption makes them conflict.
3. Look for an answer that breaks that assumption.

Often the conflict rests on a hidden causal or measurement assumption.

Common resolution patterns

• The groups are different (mixing populations).
• The measurement method changed.
• A third factor offsets the expected effect.
• The timeline matters (short-term vs long-term).

Worked example: resolving a paradox

Facts:

• “A company increased customer service staff.”
• “Customer satisfaction scores decreased.”
  A resolution might be:
• “The company changed its survey method to a stricter scoring scale at the same time.”
  This explains why scores could drop even if service improved.

Exam Focus

• Typical question patterns:
  • “Which of the following, if true, most helps to resolve the apparent discrepancy?”
  • Discrepancies based on before/after comparisons and performance metrics
  • Scientific-style results that clash with expectations
• Common mistakes:
  • Choosing an answer that strengthens one side of the discrepancy rather than reconciling both
  • Treating it like a Weaken question (attacking a conclusion) when there may be no conclusion
  • Missing that the discrepancy may be about definitions or measurement, not causality

Parallel Reasoning and Parallel Flaw: Matching Structure, Not Subject Matter

Parallel questions can feel time-consuming because they ask you to compare arguments. The key is to compare the logical skeleton rather than the topic.

Parallel reasoning

Parallel reasoning questions ask for an answer choice whose reasoning matches the stimulus.

The stimulus may be valid or invalid. Your job is not to “fix” it; it’s to match it.

Parallel flaw

Parallel flaw asks for an answer choice that makes the same mistake.

This means you must first identify the flaw in the stimulus, then find the option with the same flaw—even if the content is totally different.

What to match

Depending on the argument, match:

• Conditional structure (including contrapositive vs converse)
• Quantifiers (all/some/most)
• Causal structure (correlation-to-causation, alternative cause ignored)
• Argument form (general-to-specific, analogy, elimination)

A helpful trick: abstract the argument using letters.

Worked example: parallel flaw (affirming the consequent)

Stimulus form:

• If A → B
• B
• Therefore A

Correct answer must have that same form, even if it’s about medicine, sports, or art.

Exam Focus

• Typical question patterns:
  • Long answer choices where only one matches the logical form
  • Parallel flaw questions using conditional mistakes or quantifier shifts
  • Stimuli with multiple premises where the temptation is to match on keywords
• Common mistakes:
  • Matching based on topic similarity (“both about animals”) instead of structure
  • Missing a negation or quantifier and therefore matching the wrong form
  • Forgetting that the stimulus may be valid (and hunting for a flaw that isn’t there)

Putting It All Together: Answer Choice Strategy and Trap Patterns

Even with strong reasoning skills, LR is multiple-choice—and the LSAT is skilled at writing wrong answers that feel attractive. Your goal is to combine clear analysis with disciplined elimination.

The “job description” test

Before evaluating answers, remind yourself what the correct answer must do.

• Strengthen: must increase support.
• Necessary assumption: must be required.
• MBT: must be forced by the statements.

When an answer is tempting, ask: “Does it do the job, or is it just interesting?”

Common trap answer types

1. Out of scope: discusses a related topic but doesn’t affect the logic.
2. Too strong: uses extreme language not supported (common in MSS/MBT).
3. Too weak: technically relevant but doesn’t move the argument enough (common in Strengthen/Weaken).
4. Reversal: flips a conditional or a quantified relationship.
5. Restates a premise: feels familiar but doesn’t address the question.
6. Addresses a different conclusion: targets an intermediate conclusion or a side issue.

Calibrating strength: matching the task

A subtle but crucial skill is matching the strength of your answer to the question type.

• Sufficient assumption answers are often strong because they must guarantee the conclusion.
• MSS answers are often modest because they must be best supported without going beyond.
• Strengthen answers can vary—stronger is not always better if it introduces new controversial claims.

When (and when not) to diagram

Diagramming is most helpful when:

• The stimulus contains multiple conditional statements.
• You need to combine rules to infer something.

Diagramming is less helpful when:

• The stimulus is primarily causal, statistical, or analogy-based.

A good rule is to diagram only when it reduces complexity. If it’s faster to understand in words, stay in words.

Exam Focus

• Typical question patterns:
  • Answer choices that differ by one quantifier (“some” vs “most” vs “all”) or one logical operator (“if” vs “only if”)
  • Strengthen/weaken sets where two answers are relevant but one targets the real gap
  • Necessary assumption questions with tempting “strengtheners” that aren’t required
• Common mistakes:
  • Falling for “sounds right” answers that don’t do the stem’s job
  • Ignoring small logical words (no, some, most, only, unless)
  • Over-diagramming simple arguments and running out of time or losing the conclusion