Use this LSAT Logical Reasoning question type cheat sheet to identify what a stem requires before you judge answer choices. Each task has a different standard: some answers must follow, some only need to help, and others must describe the argument exactly.
The LR Question-Type Taxonomy
Prep companies may use slightly different labels, but these families cover the recurring LR tasks and their common variants.
| Question type | Recognizable stem signals | What the right answer must do | Compact prediction test | Common trap |
|---|---|---|---|---|
| Must Be True | “must be true,” “properly inferred” | Follow with certainty from the stimulus | Could this be false while every fact in the stimulus remains true? If yes, reject it. | Adding a plausible real-world assumption |
| Must Be False / Cannot Be True | “must be false,” “cannot be true” | Contradict what the stimulus establishes | Could this choice be true along with every fact in the stimulus? If yes, reject it. | Treating an unsupported claim as disproved |
| Most Strongly Supported | “most strongly supported,” “best supported” | Receive more support than the other choices, without needing proof | Which choice has the best textual backing and the fewest new claims? | Demanding certainty when the stem asks for best support |
| Main Conclusion | “main conclusion,” “argument’s main point” | State what the author is trying to establish | Which sentence is supported by the rest, rather than supporting something else? | Choosing an important premise |
| Method of Reasoning | “proceeds by,” “argumentative strategy” | Accurately describe the argument’s moves | Can you paraphrase the method without judging its quality? | Selecting a description of an argument the author did not make |
| Role of a Statement | “plays which role,” “functions in the argument” | Identify a sentence’s structural job | Does it provide evidence, state a conclusion, support a sub-conclusion, or give background? | Calling a premise the main conclusion |
| Flaw | “flawed because,” “vulnerable to criticism,” “error of reasoning” | Name the argument’s invalid move | What reasoning step does the author take that the evidence does not justify? | Picking a flaw that sounds familiar but is absent |
| Necessary Assumption | “assumption required,” “depends on,” “relies on” | State something the argument needs | Negate the choice. Does the argument lose its support? | Choosing a statement that would merely help |
| Sufficient Assumption | “if assumed,” “enables the conclusion,” “justifies” | Add enough support to make the conclusion follow | If this choice were true, would the premises now guarantee the conclusion? | Choosing a necessary but incomplete bridge |
| Strengthen | “most strengthens,” “most supports” | Make the conclusion more credible | Does it close a specific gap or reduce a live alternative explanation? | Restating a premise |
| Weaken | “most weakens,” “casts doubt on” | Make the conclusion less credible | Does it expose the gap, provide a counterexample, or support an alternative cause? | Attacking a detail irrelevant to the conclusion |
| Evaluate the Argument | “most useful to know,” “most helpful to evaluate” | Ask for information whose answer could matter either way | Would a yes tend to help and a no tend to hurt, or the reverse? | Choosing a fact that matters only if answered one way |
| Resolve or Explain | “most helps explain,” “apparent discrepancy,” “surprising results” | Make both sides of the puzzle possible together | Does it explain why the apparently conflicting facts can both be true? | Explaining only one fact |
| Principle | “principle most justifies,” “conforms most closely” | Match facts to a general rule, or a rule to facts | Have all conditions of the rule been met? | Choosing an appealing principle that does not fit |
| Parallel Reasoning | “most similar in its reasoning” | Match the same inferential structure, whether valid or invalid | Ignore subject matter; do the conditionals, quantities, and conclusion type match? | Matching topic or tone |
| Parallel Flaw | “most similar flawed reasoning” | Match the same invalid move | First name the original flaw, then find that exact error elsewhere. | Finding another bad argument with a different flaw |
| Point at Issue | “agree,” “disagree,” “point at issue” | State a claim on which both speakers take opposite positions | Can one speaker be marked yes and the other no? | Choosing a claim discussed by only one speaker |
EXCEPT stems reverse the selection rule: identify the underlying family first, then choose the one answer that does not perform that family’s job. “Complete the argument” variants also map to familiar tasks: a required premise is Necessary Assumption, a premise that proves the conclusion is Sufficient Assumption, and a premise that makes the conclusion more likely is Strengthen.
Start With the Task, Then Map the Argument
There is no single reading order that fits every test taker or every stimulus. The reliable habit is to know the task before committing to an answer, then identify the conclusion, evidence, and any unstated link when the question involves an argument.
For proof-based questions, keep the standard high. Must Be True answers need to follow from the text. Must Be False answers must conflict with the text. Most Strongly Supported answers are different: they can be probable rather than certain, but they must be better supported than the alternatives.
Mini-example: Must Be True versus Must Be False
Stimulus: All archivists are careful. No careful person is reckless. Mira is an archivist.
- Must Be True: “Mira is careful.”
- Must Be False: “Mira is reckless.”
A statement such as “Mira is organized” is neither established nor contradicted, so it fits neither task.
A quick conditional reminder:
- “If A, then B” allows “not B, therefore not A.”
- It does not allow “B, therefore A.”
- It does not allow “not A, therefore not B.”
That distinction often decides inference, parallel reasoning, and flaw questions.
Argument Analysis: Conclusion, Method, Role, and Point at Issue
For a Main Conclusion question, ask what the author wants you to accept. An intermediate conclusion may be supported by earlier premises and then used to support the final claim; it is not automatically the main conclusion.
Method and Role questions are structural. Translate the stimulus into plain language before reading choices: “The author cites a survey, rejects an alternative explanation, then recommends a policy.” For Role, identify the exact sentence first and ask what it does in that chain.
Point at Issue questions involve two speakers. Do not select a statement merely because it relates to the discussion. The correct answer must produce a clean split: Speaker A accepts it, and Speaker B rejects it.
Flaw and Assumption Questions: Find the Missing Link
Flaw questions ask you to describe the argument’s unsupported move. Necessary Assumption questions ask you to find a claim the argument requires. Sufficient Assumption questions ask you to supply enough information to close the gap completely.
A flaw answer need not state a missing premise. For example, an argument may conclude that a program caused an improvement merely because the improvement followed the program. The flaw is confusing sequence with causation; a necessary assumption question, by contrast, would ask what must be true for the causal conclusion to be supported.
Common flaw patterns include causal leaps, unrepresentative samples, conditional reversals, equivocation, and treating a lack of evidence as evidence of absence. Name the specific move in the stimulus before matching it to an answer choice.
Worked walkthrough: necessary assumption
Stimulus: A city claims that placing recycling bins at every curb will reduce landfill waste. The city notes that the bins will be available to every household.
Question: Which assumption does the argument require?
Conclusion: Landfill waste will decrease.
Evidence: Recycling bins will be available.
Gap: Availability matters only if people use the bins for material that otherwise would have gone to landfill.
Necessary assumption: At least some households will place in the bins material that would otherwise have been discarded as landfill waste.
Negation test: Suppose no household uses a bin for material that otherwise would have gone to landfill. The city may have placed bins everywhere, but that evidence no longer supports a reduction in landfill waste. The assumption is necessary.
A choice saying that every household will recycle every eligible item is too strong. It could help the argument, but the conclusion does not require that universal claim.
Strengthen, Weaken, and Evaluate
For Strengthen and Weaken questions, locate one precise vulnerability. The right answer need not prove or disprove the conclusion; it changes how reasonable the conclusion is.
Worked walkthrough: causal strengthening
Stimulus: A company installed free fruit in its lobby in May. In June, employee sick days declined. The company concludes that the fruit reduced employee illness.
Question: Which statement most strengthens the conclusion?
Answer pattern: No changes to remote-work rules, sick-leave reporting, or employee health benefits occurred between May and June.
Why it helps: Those changes could otherwise explain fewer recorded sick days. Ruling them out removes some alternative explanations and makes the fruit explanation more credible.
For Evaluate questions, turn each answer into a yes/no question. A good evaluation question changes the argument’s prospects in both directions.
Example: A committee proposes opening a campus study room at 7 a.m. rather than 8 a.m., arguing that students currently lack a quiet place to study early.
Useful evaluation question: Would a substantial number of students use the room between 7 a.m. and 8 a.m.?
- Yes: the earlier opening is more likely to meet a real need.
- No: the case for changing hours weakens.
Resolve, Principle, and Parallel Questions
Resolve or Explain questions are not about attacking an argument. Accept every stated fact and find information that makes the surprising combination unsurprising. If a store cut prices but earned more profit, an answer showing that sales volume rose enough to offset the lower margin could resolve the discrepancy.
Mini-example: Principle
Rule: Anyone who makes a promise must keep it unless keeping it would cause serious harm.
Facts: Dana promised to return a borrowed book, and returning it would not cause serious harm.
Application: The principle supports the conclusion that Dana must return the book. Both the promise condition and the absence of the exception are satisfied.
Parallel Reasoning and Parallel Flaw questions reward diagramming.
Mini-example: Parallel Reasoning versus Parallel Flaw
- Valid structure: All hikers carry maps. Nia is a hiker. Therefore, Nia carries a map.
- Same valid structure: All poets write. Omar is a poet. Therefore, Omar writes.
- Flawed structure: All hikers carry maps. Nia carries a map. Therefore, Nia is a hiker.
- Same flaw: All painters wear aprons. Omar wears an apron. Therefore, Omar is a painter.
Parallel Reasoning may ask for either a valid or invalid structure; match the inference itself. Parallel Flaw additionally requires that the shared structure be the same error.
For a refresher on certainty, probability, and conditional structure, try PurrLearn’s deductive vs. inductive reasoning quiz. To practice spotting alternative explanations and thinking traps that often appear in flaw questions, use the cognitive bias quiz.
Bookmarkable LR Decision Tree
- Does the stem ask what follows, must be true, must be false, or is best supported?
- Yes: use only the stimulus; distinguish proof, contradiction, and strongest available support.
- Does it ask what the author concludes, does, assumes, or gets wrong?
- Yes: map evidence, conclusion, and the relevant gap or invalid move.
- Does it ask what would help, hurt, test, or explain?
- Help: strengthen the gap.
- Hurt: weaken the gap.
- Test: find a yes/no answer that cuts both ways.
- Explain: reconcile all stated facts.
- Does it ask for a rule, a match, or a disagreement?
- Rule: test every condition and exception.
- Match: diagram structure before comparing answers.
- Disagreement: require one yes and one no.
- If two choices remain, identify the specific reason each contender fails or succeeds. During review, explain why the wrong choices were unsupported, answered the wrong task, were too strong or weak, or described the wrong argument.
FAQ
How many LSAT Logical Reasoning question types should I memorize?
Memorize the families and their standards, not every label from every prep source. The important distinction is what the answer must do: prove, contradict, best support, describe, assume, affect, explain, apply, match, or identify disagreement.
What is the difference between Must Be True and Most Strongly Supported?
Must Be True is proof-based: the answer cannot be false if the stimulus is true. Most Strongly Supported is comparative: the correct answer has the strongest support among the choices, even if it is not guaranteed.
How do I tell necessary and sufficient assumptions apart?
Negate a necessary-assumption answer. If the argument falls apart, it is necessary. A sufficient assumption is stronger: add it to the premises and the conclusion should follow completely.
What should I write in my review notes after missing an LR question?
Write the question type, the conclusion, the gap or logical structure, and one sentence explaining why your chosen answer failed. That creates an error log you can use to drill patterns rather than just repeat questions.
Conclusion
Identify the task’s proof standard first, then make every answer choice earn its place. Use the decision tree during review, and practice one family at a time until the stem signal triggers the right prediction automatically.