Logical Reasoning Skill-Builders: Analyzing Flaws, Methods, and Parallels

Flaw in the Reasoning

A Flaw in the Reasoning question asks you to diagnose what is wrong with an argument—specifically, what kind of faulty reasoning allows the author to move from the premises to the conclusion. The LSAT isn’t testing whether you agree with the conclusion in real life; it’s testing whether the conclusion is properly supported given what was said.

What it is (and what it isn’t)

An argument on Logical Reasoning has (at minimum) a conclusion (what the author is trying to prove) and premises (the support offered). A flaw is a structural defect in how the support is used. You can think of it like a bridge: the premises are the pillars, the conclusion is the roadway, and the flaw is a hidden weakness in the design. Even if the roadway could be safe in some other design, you’re judging this design.

Two important clarifications:

  1. A flawed argument can have a true conclusion. The LSAT cares about validity/strength of support, not real-world truth.
  2. A flaw is different from a missing fact. Many flawed arguments are flawed because they assume something unstated—but the flaw answer will describe the type of assumption, not a specific new fact.

Why it matters

Flaw questions train your ability to “hear” bad reasoning quickly. That skill transfers directly to:

  • Strengthen/Weaken questions (because you can predict what the argument is relying on)
  • Necessary/Sufficient Assumption questions (because you can articulate what must be true for the reasoning to work)
  • Parallel Flaw questions (because you must match one flawed pattern to another)

In other words, learning to label flaws is like learning to identify mechanical problems in a car: once you can name the failure mode, you can fix it—or find the same failure elsewhere.

How it works (a reliable approach)

A good flaw diagnosis is usually the result of a consistent routine:

  1. Find the conclusion. Ask: “What is the author trying to persuade me of?” Look for conclusion indicators (e.g., “therefore,” “thus,” “so,” “hence”)—but also be willing to find conclusions without indicator words.
  2. List the premises. What reasons are offered?
  3. Describe the reasoning jump. In plain language: “They think ____ proves ____.”
  4. Ask: what would have to be true for that jump to work? This often reveals an unstated assumption.
  5. Match to a common flaw pattern. LSAT answer choices usually describe a category of mistake.

A practical tip: when you’re stuck between two attractive flaw answers, check which one actually describes the stimulus’s core move. Wrong answers often describe a flaw that could happen in some argument about the same topic but doesn’t happen in this one.

High-frequency flaw families (with explanation)

Rather than memorizing dozens of names, it helps to understand a few “big families” of flawed reasoning the LSAT repeatedly tests.

1) Taking correlation as causation (or confusing cause and effect)

The author observes that two things occur together and concludes one caused the other—or reverses the direction.

Why it’s flawed: correlation can be explained by:

  • a third factor causing both
  • reversed causality
  • coincidence
  • a more complex causal chain
2) Unrepresentative or biased sample (generalizing from too little)

The author draws a broad conclusion about a population based on a small or biased subset.

Why it’s flawed: the premises don’t justify projecting to the whole group.

3) Part-to-whole / whole-to-part mistakes

The argument assumes what is true of a part must be true of the whole, or vice versa.

4) Necessary vs. sufficient confusion

A very common LSAT flaw is treating a requirement as if it guarantees something (or treating a guarantee as if it’s required).

  • Necessary condition: must be present, but may not be enough.
  • Sufficient condition: guarantees, but may not be required.
5) Equivocation / shifting meaning

A key term is used in two different senses, and the argument slides from one meaning to the other.

6) Attacking the person (ad hominem) or appealing to irrelevant authority

The reasoning treats a claim as false because of who said it, or true because someone “important” said it (when that authority is not relevant).

7) False dilemma / limited options

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

These “families” matter because flaw answer choices are usually written at the family level (e.g., “confuses a condition required for a condition sufficient,” “takes for granted that because two events occur together…,” “fails to consider that the sample may be unrepresentative”).

Flaw in action (worked examples)

Example 1 (causation)

Stimulus:

In the last year, the number of people who joined local gyms increased significantly. During that same year, reported cases of seasonal flu decreased. Therefore, joining a gym prevents seasonal flu.

Step 1: Conclusion: joining a gym prevents seasonal flu.

Step 2: Premises: gym membership rose; flu cases fell.

Step 3: Reasoning jump: “These moved in opposite directions, so the first caused the second.”

Flaw: Treats a correlation as establishing a causal relationship (and ignores alternative explanations—perhaps a milder flu season, better vaccination rates, changes in reporting, etc.).

Example 2 (necessary/sufficient confusion)

Stimulus:

Any city that attracts major conventions must have a large airport. Our city has a large airport, so it will attract major conventions.

Conclusion: Our city will attract major conventions.

Premise: Cities that attract major conventions must have a large airport.

Flaw: Confuses a necessary condition with a sufficient one. Having a large airport may be required to attract conventions, but it doesn’t guarantee them.

Exam Focus

  • Typical question patterns:
    • “The argument’s reasoning is flawed because it…”
    • “Which of the following most accurately describes the flaw…?”
    • “The reasoning is vulnerable to criticism on the grounds that…”
  • Common mistakes:
    • Treating the question like a fact-check—rejecting answers because they’re “not true in real life” rather than because they don’t match the argument.
    • Picking an answer that names a flaw you recognize, but not the one the stimulus actually commits.
    • Misidentifying the conclusion—if you label a premise as the conclusion, every flaw description will feel slightly off.

Parallel Reasoning

A Parallel Reasoning question asks you to find an answer choice whose argument has the same logical structure as the stimulus. The arguments can be about totally different topics; what must match is the form—how premises support the conclusion.

What it is

In a parallel reasoning task, you’re essentially doing “argument pattern matching.” The LSAT gives you a stimulus argument, then five new arguments. Your job is to pick the one built the same way.

This is not primarily about whether the answer choice is persuasive or factually plausible. It’s about whether the skeleton matches:

  • Are there the same number of moving parts?
  • Are the relationships the same (conditional, causal, quantifier-based, comparison-based)?
  • Does the conclusion follow from the premises in the same pattern?

Why it matters

Parallel reasoning questions test whether you can abstract away from content. That’s a core law-school skill: legal reasoning often turns on applying a known structure (a rule, a standard, an inference pattern) to new facts.

They also reinforce skills that help across Logical Reasoning:

  • Seeing conditional logic (if/then relationships)
  • Recognizing quantifiers (“most,” “some,” “all,” “none”)
  • Tracking negation and contrapositives
  • Distinguishing premise from intermediate conclusion

How it works (the abstraction process)

A strong approach is to translate the stimulus into a simplified “template,” then look for the same template.

Step 1: Identify conclusion and premises

Even when the structure is the key, you still need the basic map.

Step 2: Strip away topic words

Replace topic-specific nouns with letters or general labels.

Example: “All dogs are mammals” becomes “All A are B.”

Step 3: Preserve logical features

When you abstract, you must keep the parts that change the logic:

  • Quantifiers: “some” is not “most,” and “most” is not “all.”
  • Negation: “not” can flip meaning.
  • Conditional direction: “If A then B” is not “If B then A.”
  • Causal language: “causes,” “leads to,” “results in” signals a different structure than “is correlated with.”
Step 4: Check each answer choice against the template

Don’t get seduced by surface similarities. A wrong answer may mention similar subject matter or even similar vocabulary but have a different structure.

Common structural forms you’ll see

Conditional chains

Example template: If A then B. If B then C. Therefore, if A then C.

This is a valid form (a chain).

Quantifier logic

Example template: Most A are B. Most B are C. Therefore, most A are C.

This is not automatically valid; “most” doesn’t chain that way. Parallel reasoning can involve matching invalid patterns too (that overlaps with parallel flaw, discussed later).

Causal arguments

Example template: When X increased, Y decreased. So X causes a decrease in Y.

Here the form is “correlation implies causation,” which is a flaw pattern.

Parallel reasoning in action (worked example)

Example (conditional)

Stimulus:

If a company violates the regulation, it will be fined. If it is fined, its public reputation will suffer. Therefore, if a company violates the regulation, its public reputation will suffer.

Abstract template:
If A → B. If B → C. Therefore, if A → C.

Now the correct answer choice must have the same chain structure.

Correct parallel (illustrative answer choice):

If a student misses the final exam, the student will fail the course. If the student fails the course, the student will have to retake it. Therefore, if a student misses the final exam, the student will have to retake the course.

This matches exactly: If A → B; If B → C; therefore If A → C.

What a tempting wrong answer might do:

  • Reverse a conditional: If B → A
  • Switch “some” or “most” in for “all”
  • Add an extra premise or an exception
  • Reach a conclusion about B rather than about A

Exam Focus

  • Typical question patterns:
    • “Which of the following arguments is most similar in its reasoning to the argument above?”
    • “The pattern of reasoning in the argument is most closely parallel to…”
    • (Less common) “Which exhibits reasoning most similar to the flawed reasoning above?” (this drifts toward parallel flaw)
  • Common mistakes:
    • Matching by topic instead of structure—choosing the answer “about ethics” because the stimulus was about ethics.
    • Ignoring quantifiers/negations—treating “some” like “all,” or overlooking “not.”
    • Not committing to an abstract template first—without a template, every answer can feel vaguely similar.

Parallel Flaw

A Parallel Flaw question is a hybrid: you must find an answer choice whose argument commits the same type of logical mistake as the stimulus.

What it is

In parallel flaw, you aren’t matching a valid structure—you’re matching a broken structure. That means two things are simultaneously true:

  1. The correct answer will usually be flawed in a noticeable way.
  2. The specific way it is flawed must mirror the stimulus’s flaw.

A useful mindset is: “What is the engine of the mistake?” Once you can describe that engine (e.g., confusing necessary and sufficient, sampling error, correlation/causation), you can hunt for the same engine in the choices.

Why it matters

Parallel flaw questions are one of the clearest tests of whether you can name the reasoning move rather than just react to it. This is why flaw questions are a foundation: if you can’t reliably spot the flaw in the stimulus, you’ll be forced to guess among five new flawed arguments.

How it works (a two-layer matching process)

Compared to basic parallel reasoning, there’s an extra step.

  1. Diagnose the stimulus flaw (as you would in a Flaw question).
  2. Abstract the flawed pattern into a template.
  3. Scan answer choices for the same flawed pattern, paying attention to:
    • Conditional direction errors
    • Quantifier errors
    • Causal leaps
    • Illicit generalizations
    • Shifts in meaning

A major trap: many answer choices will be flawed, but in different ways. Your job isn’t to find any flawed argument; it’s to find the one flawed the same way.

Parallel flaw “signature” patterns

Here are a few especially common parallel flaw signatures.

Signature A: Affirming the consequent (conditional logic flaw)

Form:

  • If A then B.
  • B.
  • Therefore A.

This is flawed because B could happen for reasons other than A.

Signature B: Denying the antecedent (conditional logic flaw)

Form:

  • If A then B.
  • Not A.
  • Therefore not B.

Also flawed: B could still happen without A.

Signature C: Necessary/Sufficient confusion with a “must have” rule

Form:

  • All things with outcome B have requirement A.
  • This has A.
  • Therefore it has B.

That’s the same error as in the earlier airport/conventions example.

Signature D: Correlation → causation

Form:

  • When X happens, Y happens (or changes).
  • Therefore X causes Y.

Parallel flaw in action (worked example)

Example (affirming the consequent)

Stimulus:

If the lab contamination occurred, then the test results would be unreliable. The test results are unreliable. So the lab contamination occurred.

Step 1: Translate form

  • If A then B.
  • B.
  • Therefore A.

Diagnosed flaw: Affirming the consequent.

Correct parallel flaw (illustrative answer choice):

If a car’s alternator is broken, then the battery light will turn on. The battery light is on. Therefore, the alternator is broken.

Same flaw: the battery light could be on for other reasons.

Tempting wrong answers (how they differ):

  • “If A then B; not A; therefore not B” (denying the antecedent) — flawed, but not the same flaw.
  • Causal versions (“X happened before Y, so X caused Y”) — flawed, but a different engine.

A practical technique: “pre-phrase” the flaw

Before looking at choices, state the flaw in a short sentence:

  • “They treat a necessary condition as sufficient.”
  • “They conclude the cause from a mere correlation.”
  • “They assume that because a result occurred, the one sufficient condition they named must have occurred.”

That pre-phrase becomes your filter.

Exam Focus

  • Typical question patterns:
    • “Which of the following arguments exhibits flawed reasoning most similar to that in the argument above?”
    • “The pattern of flawed reasoning is most closely parallel to…”
    • “Which argument contains an error in reasoning most like the error above?”
  • Common mistakes:
    • Picking an answer with the same topic vibe (science-y, policy-y, moral-y) rather than the same logical mistake.
    • Forgetting to lock in the stimulus flaw first—if you start evaluating choices without a clear diagnosis, you’ll chase whatever seems familiar.
    • Overlooking small logical words (“only if,” “unless,” “most,” “some,” “not”) that change the flaw type.

Method of Reasoning

A Method of Reasoning question asks you to describe what the argument is doing—the strategy or logical role of its parts—rather than whether it’s good reasoning or which specific flaw it commits.

What it is

Method questions are about the argument’s procedure. Instead of “What’s wrong?” (Flaw) or “Which matches?” (Parallel), the question asks: “How does the author reason from premises to conclusion?”

This often means identifying things like:

  • Whether the author draws a general rule from examples (induction) or applies a general rule to a case (deduction-style application)
  • Whether the author uses an analogy
  • Whether the author rules out alternatives
  • Whether the author uses evidence to explain a phenomenon
  • Whether the author responds to an opposing view

Method questions can involve flawed or valid arguments—the method description can be correct even if the method is not logically sufficient.

Why it matters

Method questions build a skill that helps you across LR: recognizing argument roles. If you can quickly see “this sentence is a counterpremise,” “this is an intermediate conclusion,” “this is a general principle,” then you read faster and with more control.

Also, method descriptions are often reusable across question types. For example, if you can say “the author rules out one explanation and concludes another,” you’re better prepared for strengthen/weaken questions that attack that ruling-out step.

How it works (role-labeling and action verbs)

A reliable way to handle method questions is to translate the stimulus into a set of actions.

  1. Identify the conclusion and main support (as always).
  2. Label the role of each key claim:
    • premise/evidence
    • intermediate conclusion
    • counterargument
    • concession
    • principle/rule
    • example
  3. Summarize the author’s move using a method “verb phrase.”

Common method verb phrases include:

  • “Cites a general principle and applies it to a particular case”
  • “Uses an analogy to support a conclusion about a new situation”
  • “Dismisses an alternative explanation and infers the best remaining explanation”
  • “Generalizes from a sample to a broader population”
  • “Identifies a correlation and concludes a causal relationship”

Notice that some of these describe flawed methods (e.g., generalizing from a sample, correlation→causation) but method answers often remain neutral in tone (they may say “infers,” “concludes,” “takes for granted” depending on how evaluative the test wants to be).

Method vs. Flaw (how to keep them separate)

These two question types can feel similar because both talk about “reasoning.” The difference is the target:

  • Flaw: “What specific error makes the reasoning bad?”
  • Method: “What reasoning steps does the author take?”

Sometimes a method answer will sound like a flaw answer (e.g., “infers that because two things are correlated, one caused the other”). That’s still method—just a method description that happens to identify a bad step.

Method in action (worked examples)

Example 1 (principle applied to a case)

Stimulus:

Any policy that predictably harms the most vulnerable members of society is unjust. This proposed policy will predictably harm the most vulnerable members of society. Therefore, the proposed policy is unjust.

What the author is doing:
The author states a general principle (a rule about what makes a policy unjust) and then applies it to a particular policy to reach a conclusion.

A correct method description would sound like:

  • “Applies a general criterion to a specific case to draw a conclusion.”

A tempting wrong description might say:

  • “Generalizes from a few examples” (but there were no examples—there was a rule).
Example 2 (eliminating alternatives)

Stimulus:

The device is not malfunctioning, because if it were malfunctioning, it would fail the self-test. It passed the self-test. So the cause of the issue must be user error.

What the author is doing:
The author rules out one potential explanation (malfunctioning) using a conditional claim, then concludes another explanation (user error).

Whether that conclusion is fully justified is a separate question—the method is “eliminate one cause, infer another.” A careful student also notices a possible gap: passing the self-test might not eliminate all malfunctions, and even if malfunction is ruled out, other causes besides user error may remain.

Common traps in method answer choices

Method answers are often written with abstract language that can be hard to map to the stimulus. A few recurring traps:

  • Mismatched direction: An answer says the author “supports a generalization with a principle,” but the author actually supports a conclusion by applying a principle.
  • Wrong role: An answer claims the author “offers evidence for a claim,” when the stimulus actually assumes that claim.
  • Extra steps: An answer adds a step the author never performed (e.g., “considers and rejects an opposing argument” when no opposing view appears).

A good discipline is to demand a one-to-one match: if the answer says “rejects an alternative,” you should be able to point to the line where that alternative is raised and then rejected.

Exam Focus

  • Typical question patterns:
    • “The argument proceeds by…”
    • “Which of the following most accurately describes the method of reasoning?”
    • “The reasoning in the argument is that…”
  • Common mistakes:
    • Answering a different question—criticizing the argument (flaw) when you were asked only to describe its method.
    • Choosing an answer that is too specific to the topic; correct method answers usually work at the abstract level.
    • Missing intermediate conclusions—method questions often hinge on whether a statement is functioning as a premise or as a conclusion that then supports a further conclusion.