LSAT Logical Reasoning: Building, Analyzing, and Attacking Arguments

What an “argument” is on LSAT Logical Reasoning

On LSAT Logical Reasoning (LR), almost every question revolves around an argument—a set of statements where some statements (premises) are offered as support for another statement (the conclusion). Your job is rarely to decide whether the conclusion is “true in real life.” Instead, you evaluate whether the conclusion is well-supported given what the stimulus says.

A useful way to think about LR is: the test is measuring how well you can track support. Premises are reasons; the conclusion is what the author wants you to believe. Many wrong answers sound reasonable or factually true, but they fail because they misidentify the role of a statement, change the claim, or don’t actually affect the support relationship.

Premises vs. conclusion (and why the distinction matters)

A premise is evidence, a fact, a study result, an observation, a principle, or a rule the author treats as acceptable. A conclusion is the author’s main point—what they are trying to establish.

This distinction matters because most LR questions ask you to:

• find what must be assumed to connect premises to conclusion,
• identify what would strengthen or weaken that connection,
• describe the author’s reasoning (often by spotting a flaw), or
• determine what follows from the premises.

If you mislabel a premise as the conclusion (or vice versa), you will consistently pick tempting wrong answers.

How to find the conclusion

Often, the conclusion is signaled by conclusion indicators such as “therefore,” “thus,” “so,” “hence,” “consequently,” “it follows that,” or “clearly.” Premises are sometimes signaled by “because,” “since,” “for,” “given that,” or “in light of.”

But the LSAT frequently withholds clear indicator words, so you also need structural methods:

1. Ask: What is the author trying to prove? The conclusion is the statement that would be most controversial or most in need of support.
2. Look for support direction. Premises typically explain or justify; conclusions are explained or justified.
3. Try the “therefore” test. Insert “therefore” before candidate statements—if it sounds like the main takeaway, it’s likely the conclusion.

Subconclusions and “argument chains”

Many stimuli contain an intermediate conclusion (often called a subconclusion)—a statement that is supported by earlier premises and then used as a premise to support the main conclusion.

Being able to see chains matters because:

• Strengthen/weaken questions often target the main conclusion.
• Method/flaw questions often depend on whether the author treats a claim as established or merely supported.

Mini example (argument chain)

The sidewalks are wet, so it must have rained. And if it rained, the soccer game will be canceled. Therefore, the soccer game will be canceled.

• “The sidewalks are wet” = premise
• “It must have rained” = intermediate conclusion (supported by wet sidewalks)
• “If it rained, the soccer game will be canceled” = premise (a conditional rule)
• “The soccer game will be canceled” = main conclusion

Background information vs. premises

Some statements set context without supporting anything. LR sometimes includes background that feels like evidence but does no work. A good habit is to ask: “If I removed this sentence, would the argument lose support?” If not, it may be background.

Example: identifying structure

Stimulus:

Many residents complain that downtown traffic is worse than last year. However, the city’s traffic sensors show that average downtown travel time has decreased. Thus, the residents’ complaints are exaggerated.

Walkthrough:

• First sentence: residents complain (this is not evidence that traffic is worse; it’s evidence that people say it is).
• Second sentence: sensors show travel time decreased (a premise that undercuts the complaints).
• “Thus…” introduces the conclusion: complaints are exaggerated.

Notice the subtlety: the argument is not “traffic is better,” but “complaints are exaggerated.” That difference often drives answer choices.

Exam Focus

• Typical question patterns:
  • “Which one of the following is the conclusion?” or “The argument’s main conclusion is that…”
  • Method/flaw questions that require you to separate premises, subconclusions, and background
  • Strengthen/weaken questions where missing the main conclusion makes every answer feel irrelevant
• Common mistakes:
  • Treating an opinion report (“Critics claim…”) as the author’s conclusion
  • Confusing an intermediate conclusion for the main conclusion
  • Assuming the last sentence is always the conclusion (LSAT often places conclusions earlier)

Validity, soundness, and what the LSAT actually cares about

It’s tempting to evaluate arguments like a debate: “Is the author right?” LR is different. The core skill is evaluating whether the reasoning works, not whether the conclusion is factually correct.

Validity: does the conclusion follow from the premises?

An argument is valid (in the logical sense) if, assuming the premises are true, the conclusion must be true. Validity is about structure—whether the support is airtight.

LR contains both:

• Deductive reasoning: aims for certainty (often using conditional logic or quantifiers)
• Inductive reasoning: aims for probability (often using evidence like samples, trends, or causal claims)

The LSAT tests both, but many questions hinge on recognizing that inductive arguments can be “reasonable” without being logically guaranteed.

Soundness: valid + true premises

A sound argument is valid and has true premises. LR rarely asks you to judge truth in the real world; it asks whether the conclusion is supported given what’s stated. So you can treat premises as “given,” unless the question asks you to evaluate an assumption or flaw.

Why this matters for question types

• In Must Be True / Inference questions, you want deductive force: what follows necessarily.
• In Strengthen/Weaken questions, you often deal with inductive reasoning: what makes a conclusion more/less likely.
• In Flaw questions, you identify why the support is insufficient (often because the author treats an inductive link as if it were deductive certainty).

Example: strong vs. valid

Stimulus:

Every time the factory emits odor X, nearby residents report headaches. Therefore, odor X causes headaches.

This is not deductively valid because correlation doesn’t guarantee causation. Still, it might be a strong inductive argument if other explanations are unlikely. Strengthen/weaken questions often live in this “strong but not valid” space.

Exam Focus

• Typical question patterns:
  • Must Be True questions that test whether you can distinguish “likely” from “must”
  • Flaw questions where the author treats evidence as conclusive when it’s merely suggestive
• Common mistakes:
  • Importing real-world knowledge to reject a premise (LR usually treats premises as facts)
  • Accepting a conclusion because it seems plausible, even when it doesn’t follow
  • Confusing “supports” with “proves” in inference-style questions

Conditional reasoning (if–then): the backbone of formal LR

A large portion of LR depends on conditional reasoning—statements of the form “If A, then B.” These create rules about what must happen when a condition is met.

Anatomy of a conditional

A conditional statement has two parts:

• Sufficient condition: the “if” part (what triggers the rule)
• Necessary condition: the “then” part (what must be true if the trigger happens)

If the sufficient condition happens, the necessary condition must happen.

Example:

If a student is accepted, then the student submitted an application.

• Accepted = sufficient
• Submitted application = necessary

Contraposition (the most important conditional tool)

The contrapositive is a logically equivalent way to write a conditional. You negate and swap the terms:

If A, then B.
Contrapositive: If not B, then not A.

This is valid and guaranteed. By contrast, the following are common traps:

• Mistaken reversal: If B, then A (not equivalent)
• Mistaken negation: If not A, then not B (not equivalent)

Worked example (contrapositive)
Rule:

If the device is waterproof, then it is sealed.

Contrapositive:

If it is not sealed, then it is not waterproof.

This lets you infer things from missing necessary conditions.

“Only if,” “unless,” and other tricky wording

LR loves alternative phrasings that test whether you can correctly place sufficient vs. necessary.

Key translations:

• “A only if B” means: If A, then B. (B is necessary)
• “A if B” means: If B, then A. (B is sufficient)
• “Unless” usually introduces a necessary condition. A practical translation method is:
  • “X unless Y” often means: If not Y, then X.

Example:

The picnic will happen unless it rains.

Translation:

• If it does not rain, then the picnic will happen.
• Equivalent: If the picnic does not happen, then it rained. (contrapositive)

Linking conditionals (chains)

If you have:

• If A then B
• If B then C

You can chain them:

• If A then C

This is common in inference questions.

Conditional reasoning with necessary vs. sufficient causes

Causal language sometimes hides conditional structure.

• “A guarantees B” behaves like: If A, then B.
• “B requires A” behaves like: If B, then A.

Misreading these reverses the logic.

Example: spotting a flawed conditional inference

Stimulus:

If a company is profitable, it will invest in research. This company invested in research. So it is profitable.

This is the mistaken reversal: from “If profitable → research” the author concludes “If research → profitable.” That does not follow.

Exam Focus

• Typical question patterns:
  • Must Be True questions with several conditional rules that must be chained
  • Flaw questions involving mistaken reversal/negation
  • Strengthen questions where the correct answer supplies a missing link in a conditional chain
• Common mistakes:
  • Flipping sufficient and necessary when translating “only if,” “requires,” or “unless”
  • Forgetting that only the contrapositive is logically equivalent
  • Treating a necessary condition as if it were sufficient (“Sealed, therefore waterproof”)

Quantifiers and formal logic beyond conditionals

Not all “formal” LR is pure if–then. The LSAT also tests reasoning with quantifiers—words like “all,” “most,” “many,” “some,” “few,” and “none.” These words set how broad a claim is, and small shifts in quantifiers can completely change what follows.

Common quantifiers and what they commit you to

• All / Every: 100% inclusion (strongest)
• None: 0% inclusion
• Some: at least one (very weak; could be many)
• Most: more than half
• Many: not precise (context-dependent; weaker than “most”)

LR frequently exploits the fact that people casually treat “some” as “most,” or “most” as “all.”

What you can validly infer

From “All A are B,” you can infer:

• If A, then B (conditional form)
• If not B, then not A (contrapositive)

But you cannot infer:

• All B are A
• Some A are B (unless you also know at least one A exists)

That last point matters: universal statements don’t guarantee existence. “All unicorns are purple” doesn’t imply unicorns exist.

“Most” logic: limited but still testable

“Most” statements don’t convert cleanly into strict conditionals, but they do allow some reliable inferences.

If “Most A are B,” then:

• Some A are B (existence is implied because “most” requires a group)
• It is possible that some A are not B (in fact, it must be true unless all A are B)

But you cannot infer that a particular A is B.

Comparing strength of claims (useful for strengthen/weaken)

A recurring LR skill is recognizing that:

• A stronger claim is harder to prove but easier to weaken.
• A weaker claim is easier to defend but often doesn’t support much.

For example, “All” is easier to weaken (one counterexample) than “Most.”

Example: quantifier trap in an inference question

Stimulus:

Most of the city’s buses are electric. All electric buses require specialized maintenance. Therefore, most of the city’s buses require specialized maintenance.

This conclusion does follow: if more than half of buses are electric, and every electric bus requires specialized maintenance, then more than half require specialized maintenance.

But notice what you cannot infer:

• “All buses require specialized maintenance” (too strong)
• “This particular bus requires specialized maintenance” (unless you know it’s electric)

Exam Focus

• Typical question patterns:
  • Must Be True questions that depend on careful quantifier interpretation (especially “most”)
  • Strengthen/weaken questions where an answer subtly shifts from “some” to “most” to “all”
• Common mistakes:
  • Treating “some” as if it meant “many” or “most”
  • Assuming universal statements imply existence
  • Missing that one counterexample destroys an “all” claim but not a “most” claim

Causal reasoning: how LR builds and attacks cause-and-effect claims

LR loves causal arguments because they’re intuitive—and because they’re easy to do poorly. A causal claim says that one thing (the cause) produces or influences another (the effect). These arguments are common in strengthen, weaken, flaw, and evaluate question types.

The basic causal form

A typical causal argument looks like:

• Premise: When X happens, Y happens.
• Conclusion: Therefore, X causes Y.

The reasoning gap is that “X and Y occur together” doesn’t prove “X produces Y.” The LSAT tests whether you can spot alternative explanations.

The four classic ways causal arguments go wrong

1. Alternative cause: Something else causes Y, and X just happens to be present.
2. Reverse causation: Y causes X rather than X causing Y.
3. Common cause: A third factor causes both X and Y.
4. Coincidence / bad data: The association is accidental or based on unrepresentative data.

Strengthening a causal claim

To strengthen “X causes Y,” you typically want to:

• rule out alternative causes,
• show the effect happens when the cause is present and not when it’s absent,
• establish a plausible mechanism,
• show the relationship holds across settings (replication).

Weakening a causal claim

To weaken, you typically:

• provide an alternative explanation,
• show cases where X occurs without Y (or Y without X),
• point out confounds in the data.

Causation vs. correlation: what the LSAT expects

The LSAT doesn’t require scientific expertise; it tests reasoning patterns. You don’t need to know biology to see that a study could be confounded. The core is: association is not enough.

Worked example: weakening a causal argument

Stimulus:

After the city installed brighter streetlights, nighttime crime decreased. Therefore, brighter streetlights reduce crime.

How to think:

• Evidence: a before/after trend.
• Hidden assumptions: nothing else relevant changed; the trend isn’t seasonal; crime wasn’t displaced to other areas.

A strong weaken answer might say:

During the same period, the city increased police patrols in the same neighborhoods.

This introduces an alternative cause.

Evaluate-the-argument questions and causal reasoning

In Evaluate questions, you’re asked what information would be most helpful to determine whether the argument works. For causation, evaluation often centers on:

• whether other factors changed,
• whether the relationship holds when you control variables,
• whether the supposed cause precedes the effect.

Exam Focus

• Typical question patterns:
  • “Which answer, if true, most strengthens/weaken the argument?” with a correlation-to-causation leap
  • “Which question would be most useful to ask to evaluate the argument?” targeting confounds
  • Flaw questions describing confusion of correlation and causation
• Common mistakes:
  • Weakening with something irrelevant (true but not an alternative explanation)
  • Strengthening by restating the conclusion instead of adding support
  • Missing reverse causation when the timeline is unclear

Assumptions: the hidden bridge between premises and conclusion

An assumption is an unstated idea that the argument requires (or relies on) for the premises to support the conclusion. Assumption questions are central in LR because they test whether you can see what the author took for granted.

Why assumptions exist

Arguments are usually incomplete. Authors omit steps because they seem obvious, because of limited space, or because they’re persuading rather than proving. The LSAT exploits this by asking you to identify what must be true for the reasoning to work.

A helpful metaphor: premises are one side of a river, conclusion is the other. The assumption is the bridge. If the bridge collapses, the conclusion no longer follows.

Necessary vs. sufficient assumptions

These are different, and LR tests both.

• A necessary assumption is something the argument must rely on. If it’s false, the argument falls apart.
• A sufficient assumption is something that, if added, guarantees the conclusion (it makes the argument valid).

A necessary assumption does not have to prove the conclusion by itself; it just has to be required.

How to find the necessary assumption

A reliable method is the negation test:

1. Take an answer choice.
2. Negate it (logically, not just by adding “not”).
3. Ask whether the negated statement would destroy the argument.

If negating it seriously undermines the reasoning, it’s likely necessary.

Negation examples (common patterns)

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

How to find the sufficient assumption

A sufficient assumption is often stronger. You’re looking for a statement that, when combined with the premises, forces the conclusion.

A practical approach:

• Identify the gap: what link is missing?
• Look for an answer that explicitly connects the key premise idea to the conclusion idea—often by using conditional language or a universal claim.

Table: necessary vs. sufficient assumption (quick conceptual comparison)

Worked example: necessary assumption

Stimulus:

The museum should extend its hours. When the museum stays open later, more people visit. Therefore, extending hours will increase ticket revenue.

Structure:

• Premise: Later hours → more visitors.
• Conclusion: Extending hours → more revenue.

Gap: more visitors does not necessarily mean more revenue. Maybe those visitors enter free.

A necessary assumption is something like:

• Visitors who come during extended hours will pay for tickets (or at least won’t reduce average revenue per visitor so much that revenue doesn’t increase).

Notice: the necessary assumption doesn’t have to claim revenue will increase; it just has to block the possibility that more visitors yields no revenue increase.

Common assumption traps

• Out of scope: an answer about a side issue the argument doesn’t depend on.
• Too strong (for necessary): claims more than required.
• Restating a premise: something already said can’t be the missing assumption.

Exam Focus

• Typical question patterns:
  • “The argument requires assuming which of the following?” (necessary)
  • “Which of the following, if assumed, allows the conclusion to be properly drawn?” (sufficient)
  • Strengthen questions that function like “find the missing assumption”
• Common mistakes:
  • Using the negation test incorrectly (negating too loosely or not logically)
  • Confusing necessary with sufficient and choosing an answer that merely helps a bit
  • Picking an answer that strengthens but is not required (common in necessary assumption questions)

Common argument flaws (and how the LSAT describes them)

A flaw is a recurring pattern of bad reasoning—an unreliable move from premises to conclusion. Flaw questions ask you to match the stimulus’s mistake to a description in the answer choices.

Why flaw recognition is so valuable

Even when a question is not explicitly a “Flaw” question, recognizing flaws helps you:

• predict how to weaken or strengthen,
• identify assumptions,
• avoid attractive wrong answers that repeat the flawed idea.

Major flaw families

1) Confusing correlation with causation

As discussed earlier, “X and Y occur together, so X causes Y.”

2) Mistaken necessary/sufficient (conditional confusion)

Common forms:

• “If A then B; B; therefore A” (mistaken reversal)
• “If A then B; not A; therefore not B” (mistaken negation)

3) Sampling and generalization flaws

These occur when an argument draws a broad conclusion from a non-representative or too-small sample.

Typical LSAT versions:

• Survey of website visitors used to infer what the whole population thinks
• Study done on a narrow group used to infer universal human behavior

To evaluate these, ask:

• Who was sampled?
• Is that group representative of the target group?
• Could selection bias be present?

4) Equivocation (shifting meaning)

The argument uses a key word in two different senses.

Example:

A “theory” is just a guess. Evolution is a theory. So evolution is just a guess.

Here “theory” shifts from scientific sense (well-supported explanatory framework) to casual sense (speculation). You don’t need science knowledge—the flaw is the shift in meaning.

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

• Composition: what’s true of the parts is assumed true of the whole.
• Division: what’s true of the whole is assumed true of each part.

Example (composition):

Each ingredient in this dish is inexpensive, so the dish must be inexpensive.

Not necessarily—preparation costs may be high.

6) False dilemma (limited alternatives)

The argument assumes only two options exist when more are possible.

Example:

Either we ban cars downtown or traffic will only get worse.

Maybe there are other traffic solutions.

7) Circular reasoning

The conclusion is effectively assumed in the premise.

Example:

This policy is unjust because it is unfair.

“Unjust” and “unfair” are essentially the same claim.

Worked example: diagnosing a flaw

Stimulus:

The new tutoring program must be effective. After it was introduced, average grades increased.

Flaw: assumes the program caused grade increases; ignores other changes (curriculum, grading policy, selection effects).

A correct flaw description might say:

• “Takes for granted that because one event followed another, the first caused the second.”

Exam Focus

• Typical question patterns:
  • “The reasoning is flawed because the argument…” followed by abstract descriptions
  • Parallel flaw questions asking you to find the same mistake in a new stimulus
• Common mistakes:
  • Choosing an answer that criticizes the argument but not in the way it is actually flawed
  • Missing subtle conditional errors because the language isn’t in obvious if–then form
  • Confusing “a premise is false” with “the reasoning is invalid” (LSAT focuses on the reasoning)

Strengthen, weaken, and evaluate: manipulating support

These question types test your ability to change how likely a conclusion is, given the premises. The key is that the correct answer is the one that most affects the premise-to-conclusion link, not the topic in general.

Strengthen questions: what “strengthen” really means

To strengthen is to make the conclusion more likely to follow from the premises. The best strengthen answers often:

• support an assumption the argument relies on,
• eliminate an alternative explanation,
• provide additional evidence that points in the same direction,
• show the reasoning pattern is reliable.

A crucial mindset: you are not trying to find an answer that proves the conclusion; you’re looking for the one that best improves support.

Weaken questions: targeting the link

To weaken is to make the conclusion less likely, often by:

• providing counterevidence,
• revealing an alternative explanation,
• showing the premise is irrelevant or unreliable,
• pointing out missing cases where the premise doesn’t lead to the conclusion.

Strong weaken answers frequently attack an assumption.

Evaluate questions: what information would matter most?

Evaluate questions ask for a fact (or question) that would help decide whether the argument is good. The best evaluate choice usually identifies a fork in the road: if the answer goes one way, the conclusion is supported; if it goes the other way, the argument collapses.

A practical way to spot the right evaluate answer:

• Identify the conclusion.
• Ask: “What would I want to know to judge whether the premises truly support that conclusion?”

Worked example: strengthen vs. weaken vs. evaluate

Stimulus:

People who drink green tea have lower rates of colds. Therefore, drinking green tea prevents colds.

• Strengthen: “In a randomized experiment, participants assigned to drink green tea had fewer colds than those assigned to drink a placebo beverage.”
• Weaken: “People who drink green tea also tend to get more sleep and exercise more.”
• Evaluate: “Were the green tea drinkers and non-drinkers similar in other health-related behaviors?”

Notice how each correct move targets the correlation-to-causation gap.

Exam Focus

• Typical question patterns:
  • Strengthen/weaken with causal claims, surveys, or policy recommendations
  • Evaluate questions that ask which piece of information would be most helpful to assess the argument
• Common mistakes:
  • Picking an answer that is merely on-topic but doesn’t affect the reasoning
  • Confusing “weaken the conclusion” with “introduce a negative fact” (it must undermine support)
  • In evaluate questions, choosing a question whose answer wouldn’t change your confidence either way

Inference and “Must Be True”: drawing what follows from the text

Inference-style questions go by several names (e.g., Must Be True, Most Strongly Supported, Soft Must Be True). The core is: you are given premises, and you must pick an answer that is supported by them.

The strictness spectrum

• Must Be True: the correct answer must follow; anything that could be false is wrong.
• Most Strongly Supported: slightly looser; the correct answer is best supported, even if not 100% provable.

In practice, both reward the same habit: stay disciplined and avoid adding assumptions.

How inference differs from argument evaluation

In inference questions, the stimulus may not even be an argument; it can be a set of facts. You’re not asked to critique reasoning—you’re asked to derive.

A reliable inference process

1. Treat statements as constraints: imagine each sentence narrowing the possible world.
2. Combine statements carefully: look for overlaps (shared groups, conditional chains, quantifiers).
3. Be wary of attractive overstatements: wrong answers often go one step too far.

Worked example: conditional inference

Stimulus:

All employees who work remotely must log their hours daily. Some employees work remotely.

What follows?

• At least some employees must log their hours daily.

What does not follow?

• All employees must log their hours daily.
• Some employees do not log their hours daily.

The stimulus tells you remote workers must log hours, and some workers are remote—so at least one person must log hours.

Worked example: combining quantifiers

Stimulus:

Most of the committee members are lawyers. No lawyers on the committee are engineers.

Inference:

• Most of the committee members are not engineers.

Because more than half are lawyers, and lawyers are a subset of “not engineers,” more than half are not engineers.

Exam Focus

• Typical question patterns:
  • “Which of the following can be properly inferred?” using conditional chains and quantifiers
  • “Most strongly supported” where several answers are plausible but only one is consistently supported
• Common mistakes:
  • Treating “could be true” as “must be true”
  • Adding outside information or assumptions to make an answer work
  • Falling for answers that repeat stimulus language but subtly strengthen the claim

Method of reasoning, role, and point-at-issue: reading like a logician

Not all LR questions are about adding information (strengthen/weaken) or deriving consequences (inference). Many are about describing what the author did.

Method of reasoning

A Method of Reasoning question asks you to describe the argument’s strategy in abstract terms—what the author’s reasoning does.

Common methods include:

• offering an explanation for a phenomenon,
• drawing a general principle from examples,
• applying a general principle to a specific case,
• rejecting an opposing view and replacing it with a different explanation,
• using an analogy.

To answer, you must first have a clean premise/conclusion breakdown. Then you match the description.

Example (method)
Stimulus:

Some people claim the new law will reduce pollution. But pollution levels did not decrease after similar laws were passed elsewhere. So the new law will not reduce pollution.

Method: challenges a prediction by citing evidence from similar cases.

Role of a statement

Role questions ask what a particular sentence does in the argument: premise, conclusion, intermediate conclusion, background, counterpremise, etc.

A frequent twist: the stimulus includes a counterpremise—a statement the author mentions but does not endorse, often introduced with “some argue,” “critics say,” or “it might be thought.” The role of that statement is typically “a claim the author disputes.”

Point at issue

In Point at Issue questions, two speakers disagree. Your job is to identify a statement that one would agree with and the other would not (or vice versa).

The key is to track:

• each speaker’s conclusion,
• any shared premises,
• where they diverge (often in assumptions or causal interpretations).

A good technique is to paraphrase each speaker’s bottom line in one sentence before going to the answers.

Exam Focus

• Typical question patterns:
  • “The argument proceeds by…” (method) with abstract answer choices
  • “The statement ‘X’ plays which role?” (role)
  • Two-speaker dialogues asking what they disagree about (point at issue)
• Common mistakes:
  • Answering method questions with a topic summary instead of a reasoning description
  • Mislabeling a counterpremise as the author’s view
  • In point-at-issue, choosing a statement both speakers would reject (no disagreement)

Parallel reasoning, parallel flaw, and principle questions

These question types test whether you can recognize the structure of reasoning independent of topic.

Parallel reasoning

A Parallel Reasoning question asks you to find an argument that uses the same logical pattern.

The most efficient approach is to abstract the stimulus:

• Replace content with variables (A, B, C).
• Identify whether it’s conditional, causal, quantifier-based, or analogical.
• Note whether it’s valid or flawed.

If the original is flawed, the match must be flawed in the same way.

Parallel flaw

Parallel Flaw is the same idea, but explicitly about matching the mistake.

A common trap is choosing an answer that has a similar topic or tone but a different logical error.

Principle questions

Principle questions come in two main forms:

1. Principle—Strengthen/Justify: choose a general rule that, if true, supports the argument.
2. Principle—Conform: choose a general rule that the argument follows (a principle the author’s reasoning exemplifies).

These reward the same skill: matching abstract structure.

Worked example (principle justify)
Stimulus:

The editor should print the retraction, because the article contained a factual error.

A justifying principle might be:

Whenever a publication prints a factual error, it should print a retraction.

Notice how the principle bridges premise to conclusion in a rule-like way.

How to avoid being fooled by surface similarity

Parallel and principle questions are designed so that several answers “sound similar.” You must prioritize:

• logical form (conditional direction, quantifier strength),
• whether the conclusion is broader than the premises,
• whether the argument is valid or commits a specific flaw.

Exam Focus

• Typical question patterns:
  • “Which argument is most similar in its reasoning?” (parallel)
  • “Which exhibits an error most similar to…” (parallel flaw)
  • “Which principle, if valid, most helps to justify the conclusion?” (principle justify)
• Common mistakes:
  • Matching on topic instead of structure
  • Missing that the original argument is flawed (and choosing a valid parallel)
  • In principle questions, picking a principle that is related but too weak to support the conclusion

Resolving paradoxes, explaining discrepancies, and reasoning with competing facts

Some LR stimuli present a situation that seems inconsistent—two facts that “shouldn’t” both be true. These often appear in Resolve the Paradox (also called Explain the Discrepancy) questions.

What these questions are really asking

You are not trying to strengthen or weaken an argument. You are trying to find an answer that, if true, makes the surprising set of facts make sense.

The right answer typically:

• introduces a distinction (different groups, different time periods, different definitions),
• identifies a hidden factor that reconciles the facts,
• shows that what looked like a conflict is not actually a conflict.

The “two facts” framework

Treat the stimulus as containing:

• Fact A
• Fact B
• Surprise: A and B seem incompatible.

Your job is to find a statement C such that A + B + C can all be true together.

Worked example: resolving a discrepancy

Stimulus:

A restaurant lowered its prices, expecting more customers. However, after lowering prices, its total revenue fell.

A resolution might be:

After the price reduction, the restaurant served fewer customers because construction blocked access to the street.

This allows lower prices and lower revenue to coexist by explaining why the expected customer increase didn’t happen.

Other possible resolutions could involve:

• customers buying less per visit,
• higher costs reducing net revenue (if the stimulus used profit vs. revenue carefully),
• discounts applying only to low-margin items.

Common wrong-answer patterns

• Answers that explain only one fact, not the discrepancy.
• Answers that introduce a new contradiction.
• Answers that strengthen the “surprise” rather than resolve it.

Exam Focus

• Typical question patterns:
  • “Which of the following, if true, most helps to resolve the apparent discrepancy?”
  • Stimuli with before/after changes, conflicting study results, or surprising outcomes
• Common mistakes:
  • Treating it like a weaken question and picking an answer that attacks one side
  • Ignoring key words like revenue vs. profit, average vs. total, percentage vs. absolute number
  • Choosing an answer that could be true but doesn’t actually make the two facts compatible

Efficient LR habits: predicting gaps, controlling scope, and answer-choice discipline

LR rewards careful reading, but it also rewards consistency—using the same mental moves across many stimuli.

Prephrasing: predicting what you need before looking at answers

Prephrasing means you briefly articulate what would solve the problem before reading choices.

• In strengthen/weaken: identify the gap and predict the type of information that would affect it.
• In assumption: name the missing bridge.
• In resolve: state why the facts conflict and what kind of fact would reconcile them.

You won’t always predict the exact right answer, but prephrasing prevents you from being led by tempting wording.

Scope control: don’t fight the wrong battle

Many wrong answers are:

• too broad (addressing a bigger claim than the conclusion), or
• too narrow (addressing a detail unrelated to the reasoning).

A good question to ask is: “Does this answer interact with the conclusion’s claim and the premises’ support?” If it doesn’t, it’s probably wrong.

Degree words and logical force

LR is extremely sensitive to words like:

• all/none vs. some
• must vs. may
• always vs. sometimes
• only/mainly/primarily

Wrong answers often change the force just enough to be unsupported.

The “why this answer, not that one” mindset

When stuck between two choices, compare them by asking:

• Which one more directly addresses the argument’s central gap?
• Which one introduces fewer new assumptions?
• Which one matches the required strength (must vs. likely; necessary vs. sufficient)?

This keeps you from picking an answer because it “sounds smart.”

Exam Focus

• Typical question patterns:
  • Any LR question where multiple choices seem relevant but only one precisely targets the logical gap
  • Assumption questions where two answers strengthen but only one is required
• Common mistakes:
  • Letting answer choices redefine the issue (scope shift)
  • Ignoring quantifier/degree changes (some vs. most; can vs. will)
  • Failing to commit to an argument core (premises + conclusion) before evaluating answers