LSAT flaw questions ask you to name the error in an argument. You don't fix the argument. You find the jump it makes, then pick the answer that calls that jump a mistake. Below you'll find the four-step method I teach, two original examples, and the traps that make wrong answers look right.
How to spot LSAT flaw questions
A flaw answer names this argument's jump as an error. Say the jump in plain words first, then translate it into the wording the LSAT uses. The question stem asks what is wrong with the reasoning, usually with words like "flawed" or "vulnerable to criticism."
Start by asking where the argument goes wrong. There are three families of flaws, and a few words in the stimulus point you to the right one.
| If the stimulus says | Family | What it covers | Your tool |
|---|---|---|---|
| "if," "only," "all," "some" | Formal | Reversal, negation, circular reasoning, false choice | Diagram it |
| "study," "survey," "data," "example" | Evidence | Sample, cause, analogy, part-whole, missing evidence | Name the jump |
| A key word used twice, or a person attacked | Language and source | Shifting words, straw man, attacking the speaker | Check each word |
The four-step method
- Find the conclusion and the givens. What is the argument trying to prove, and what does it offer as support?
- Name the jump in plain words. Do this before you read the answer choices.
- Translate it into flaw language. Say it simply, then find the fancy version.
- Check every part against this argument. If any piece of an answer didn't happen, cross it out.
Flaw answers reuse a small set of phrases. Here's how to decode them:
| The answer says | It means |
|---|---|
| "takes for granted" | assumes without support |
| "fails to consider" | ignores a possibility |
| "confuses X with Y" | treats two things as one |
And here is how two plain-word thoughts turn into answer language:
| You think | The answer says |
|---|---|
| "They only asked gym members." | "relies on a sample that may be unrepresentative" |
| "Needed isn't the same as enough." | "mistakes a necessary condition for a sufficient one" |
Learn the shapes, not the wording
Sixteen errors cover nearly every flaw question. Each has a shape you can recognize and a standard way the LSAT words it.
| Flaw | What it looks like | How the LSAT says it |
|---|---|---|
| Necessary vs. sufficient | Rain means wet. It's wet, so it rained. | mistakes a necessary condition for a sufficient one |
| Correlation to cause | Ice cream and sunburns rise together. | confuses correlation with causation |
| Bad sample | Polled gym members about the whole town. | relies on an unrepresentative sample |
| Part to whole | Each brick is light, so the wall is. | assumes the whole shares its parts' traits |
| Bad analogy | It worked for A, so it will for B. | overlooks a relevant difference |
| No evidence = false | No proof it's safe, so it's harmful. | treats lack of evidence as proof it's false |
| Possible to certain | It could rain, so it will. | treats what is possible as certain |
| Circular | He's honest; he says he never lies. | assumes what it sets out to prove |
| False choice | Ban cars or accept dirty air. | ignores options other than the two given |
| Shifting word | "Fair" means two things midway. | uses a key term in two different senses |
| Attack the source | She's paid by them, so she's wrong. | attacks the person, not the argument |
| Straw man | Bends the opponent's view, then beats it. | distorts the position it rejects |
| Popularity / authority | Most people believe it, so it's true. | takes popularity or status as proof |
| Number vs. rate | More crashes in City A, so it's riskier. | confuses a total with a percentage |
| Belief to truth | Everyone expects it, so it will happen. | treats a belief as if it were a fact |
| Overreach | One study hints at it, so it's proven. | concludes more than the evidence allows |
If "needed" versus "enough" still feels slippery, the LSAT necessary assumption guide is a good companion read. It's all about what an argument needs.
Worked example 1: a bad sample
Fictional argument, written for this guide.
The Harbor Hill Gazette surveyed readers leaving the Saturday farmers' market, asking whether Main Street should close to cars on weekends. Nearly every reader said yes. The Gazette concludes that most Harbor Hill residents want the street closed.
Step 1. The conclusion is that most residents want the street closed. The given is that market-goers said yes.
Step 2. In plain words: "They only asked people at the market." Those people may be the ones most likely to enjoy a car-free street.
Step 3. The word "surveyed" points to the evidence family, and the jump is a sample. The LSAT version is "relies on a sample that may be unrepresentative."
Step 4. Check the choices, shortened here:
- (A) relies on a sample that may be unrepresentative of the town's residents. Correct. Every part of it happened.
- (B) confuses correlation with causation. Nobody claimed a cause.
- (C) reports a survey result and then states a conclusion about what residents want. This only describes the argument. It never says that the move is a mistake.
- (D) assumes that every resident supports the closure. The argument said "most," not "every."
Worked example 2: needed isn't enough
Fictional argument, written for this guide.
Orion Bakery's sourdough sells out only on days when the flour delivery arrives before opening. The flour delivery arrived before opening today. So the sourdough will sell out today.
Step 1. The conclusion is that the sourdough will sell out today. The givens are the "only" rule and today's early delivery.
Step 2. In plain words: "Needed isn't the same as enough." An early delivery is required for a sellout. It doesn't guarantee one.
Step 3. The word "only" points to the formal family, so diagram it. A sellout requires an early delivery, so the diagram reads Sellout → Early delivery. The argument runs it backward: Early delivery → Sellout. The LSAT version is "mistakes a necessary condition for a sufficient one."
Step 4. Check the choices, shortened here:
- (A) mistakes a necessary condition for a sufficient one. Correct.
- (B) uses a key term in two different senses. The words never shift.
- (C) moves from a claim about sellouts to a prediction about one particular day. That describes a step. It doesn't call the step an error.
- (D) contradicts itself. The argument is flawed, but it never contradicts itself.
Traps that look right
Name the trap, then skip it fast. Wrong answers keep falling into the same three.
Real flaw, wrong argument
The answer names a genuine error that this argument doesn't commit. (You may also hear it called "right flaw, wrong argument.") Choice (B) in each example is this trap. The error is real in general, but the argument in front of you never makes it.
Describes, doesn't criticize
The answer is an accurate summary of a move that isn't a mistake. A close cousin is an answer that accurately describes a premise. It's true, so it feels safe, but a flaw answer has to name an error, not a step. Choice (C) in each example is this trap.
Too strong
The answer charges the argument with more than it did. "Contradicts itself" and "proves the opposite" are the usual suspects when the argument does neither. Choice (D) in each example is this trap.
When you miss a flaw question, write down which trap pulled you in. That one line is worth adding to your LSAT blind review.
Before you answer
- The jump is named before you read the choices.
- Every part of the answer actually happened in the argument.
- The answer names an error, not just a step.
Next step
Small live classes, capped at 12 students. Alan teaches every session himself and reviews your mistakes.
Related guides
- LSAT Necessary Assumption: How to Use the Negation Test
A necessary assumption is what the argument can't live without. Find the gap, predict something modest, and negate each finalist to check it.
- LSAT Blind Review: How to Do It and Log Your Errors
Re-solve before you check the key, then log why each miss happened. Here is the cycle, with two worked examples.