Logical Reasoning Question Types & Identification

What You Need to Know

Logical Reasoning (LR) question types are repeatable patterns. If you can identify the type quickly, you’ll know:

  • What counts as “proof” (e.g., must be true vs merely helpful)
  • Where the answer must come from (strictly the stimulus vs adding a missing link)
  • How strong the correct answer should sound (certain vs probabilistic)

The core rule

Your job is defined by the question stem. The same stimulus can support multiple tasks, but the stem tells you which “lens” to use.

The big split: Argument vs Fact Set

  1. Argument stimuli (there is a conclusion supported by premises)
    • Common tasks: Strengthen, Weaken, Assumption, Flaw, Method, Parallel, Principle
  2. Fact-set / information stimuli (no conclusion—just claims)
    • Common tasks: Must Be True / Inference, Most Strongly Supported
  3. Phenomenon / discrepancy stimuli (facts that don’t fit neatly)
    • Common task: Resolve / Explain the discrepancy

Critical reminder: You can’t reliably identify question type by stimulus alone. Read the stem early (often first) to avoid solving the wrong problem.

Step-by-Step Breakdown

Use this every time to classify the question in under 10 seconds.

  1. Read the question stem first (or immediately after a quick glance).

    • Look for “if true,” “requires,” “most strongly supported,” “flaw,” “principle,” “except,” “disagree,” etc.
  2. Flag any “stem modifiers” that change the task.

    • EXCEPT / NOT / LEAST / FALSE = you’re hunting for the odd one out.
    • IF TRUE = you’re choosing an answer that does something (strengthen/weaken/resolve).
    • MOST (most strongly supported) = you’re allowed a weaker conclusion than “must.”
  3. Classify the stimulus quickly.

    • Find the conclusion (if any). Look for indicator words (see tables below).
    • If no conclusion: treat it as fact-set unless it’s a discrepancy.
  4. Match the stem to a question-type “goal.”

    • Inference: “What follows?”
    • Strengthen/Weaken: “What changes the support?”
    • Assumption: “What must be true / what completes the proof?”
    • Flaw/Method: “Describe what the argument did wrong / how it reasoned.”
  5. Choose the right standard of proof.

    • Must be true / required / depends on → near-certain, airtight.
    • Most strongly supported → best-supported, even if not guaranteed.
    • Strengthen/Weaken/Resolve → effect-based: does it move the needle in the right direction?

Mini worked identification examples (fast)

Example A (Inference)

  • Stem: “Which of the following is most strongly supported by the statements above?”
  • ID: Most Strongly Supported (MSS) (fact-set inference, softer than Must Be True)
  • What you do: pick the choice that is best backed by the given info (often a careful restatement/combination).

Example B (Necessary Assumption)

  • Stem: “The argument depends on which of the following assumptions?”
  • ID: Necessary Assumption (NA)
  • What you do: find what must be true for the reasoning to work; typically bridges a gap.

Example C (Resolve a discrepancy)

  • Stimulus: two facts that seem inconsistent.
  • Stem: “Which of the following, if true, most helps to resolve the apparent discrepancy?”
  • ID: Resolve/Explain (Paradox)
  • What you do: look for a fact that makes both facts compatible (often by adding a distinction, time shift, definition shift, or hidden factor).

Key Formulas, Rules & Facts

Quick stem-to-type map (high-yield)

Question TypeCommon stem language (signals)Your goalProof standard / notes
Must Be True (MBT) / Inferencemust be true,” “properly inferred,” “follows logically”Derive a statement guaranteed by the stimulusStrict: only what’s forced by the text
Most Strongly Supported (MSS)most strongly supported,” “best supported”Choose the best-backed optionLess strict than MBT; avoid extreme certainty unless justified
Main Conclusion“main conclusion,” “conclusion of the argument”Identify what the premises supportOften one sentence; watch for intermediate conclusions
Method of Reasoning“argument proceeds by,” “reasoning technique,” “method”Describe the structure of reasoningAbstract the argument; match form, not topic
Flaw“flaw,” “error,” “reasoning is vulnerable to criticism”Identify what’s wrongMany answers are classic patterns (sampling, causation, equivocation, etc.)
Strengthenif true, most strengthens,” “provides the most support”Increase support for conclusionDoesn’t need to prove; just help most
Weakenif true, most seriously weakens,” “casts most doubt”Reduce supportLook for alternative causes, counterexamples, attacks on link
Necessary Assumption (NA)assumption required,” “depends on,” “presupposes,” “relies on”Find what must be trueUse Negation Test (see below)
Sufficient Assumption (SA) / Justify“assumption that enables,” “allows conclusion to be properly drawn,” “if assumed, conclusion follows”Add a premise that makes argument validOften stronger than NA; may “bridge” fully
Evaluate (the argument)“which would be most useful to know,” “helps to evaluate”Identify a question/fact that would matterUsually framed as yes/no or a key missing fact
Resolve / Explain (Paradox)“resolve the discrepancy,” “explain,” “reconcile,” “apparent contradiction”Make both facts true togetherNot weakening or strengthening—harmonizing
Principle (Strengthen)“principle that supports/justifies,” “most helps to justify”Select a general rule that backs the reasoningOften conditional-sounding; matches the argument’s logic
Principle (Conform)“argument conforms to which principle,” “most closely follows”Pick a rule the argument exemplifiesYou’re matching what they did, not what they should do
Parallel Reasoning“most similar reasoning,” “most closely parallels”Match logical formIgnore topic; track quantifiers, conditionals, causation, etc.
Parallel Flaw“contains flawed reasoning most similar to”Match the same kind of mistakeMust share flaw type + structure
Point at Issue / Disagree“disagree about,” “point at issue,” “would disagree”Find statement one says yes, other says noMust be binary: one clearly agrees, other clearly disagrees
Role of a Statement“plays which role,” “functions as”Identify what a sentence doesPremise, conclusion, subsidiary conclusion, background, objection, etc.
Complete the Argument“most logically completes,” “the argument can be best completed by”Add a missing conclusion/premiseTreat like a gap-fill; rely on what’s already there
Cannot Be True“cannot be true,” “must be false”Find the impossible optionEquivalent to MBT but inverted: find contradiction with stimulus

Conclusion & premise indicator words (for fast stimulus classification)

TypeCommon indicatorsWhat to do
Conclusion indicatorstherefore, thus, hence, so, consequently, clearly, shows that, proves that, indicates thatThe statement after/before is likely the conclusion
Premise indicatorsbecause, since, for, given that, as shown by, after allThe statement is support
Contrast / counterpremise markersbut, however, yet, although, nevertheless, despiteOften signals the real conclusion nearby

The “EXCEPT / LEAST” inversion rule

If the stem says EXCEPT, NOT, LEAST, or FALSE, you are usually looking for the one answer that does NOT fit.

  • For Strengthen EXCEPT: four strengthen, one doesn’t.
  • For MBT EXCEPT: four are supported, one isn’t.

Warning: With “EXCEPT” stems, don’t “prove” an answer right—prove four answers fit, then pick the leftover.

Standards of proof (how strong should the correct answer sound?)

  • MBT / NA: strong necessity language is okay only if forced (“must,” “cannot,” “requires”).
  • MSS: moderate language often wins (“likely,” “some,” “often”) unless stimulus warrants certainty.
  • Strengthen/Weaken/Resolve: focus on impact, not certainty.

Examples & Applications

Below are representative identification drills—what the question is asking and what you should be looking for.

1) Must Be True vs Most Strongly Supported

Stem A: “Which of the following statements is properly inferred from the information above?”

  • Type: MBT
  • Key insight: Only choose something that must follow; avoid any choice that adds a new idea.

Stem B: “Which of the following is most strongly supported by the statements above?”

  • Type: MSS
  • Key insight: You can pick a choice that’s highly plausible from the facts, even if not deductively certain.

How it appears as a trap: LSAT will give you an MSS question where one answer sounds like an MBT (“must,” “all,” “never”). Usually wrong unless the stimulus is that strong.

2) Necessary Assumption vs Sufficient Assumption

Stem A: “The argument depends on which assumption?”

  • Type: Necessary Assumption (NA)
  • Key insight: The correct answer is something the author must believe; negating it wrecks the argument.

Stem B: “Which assumption, if made, allows the conclusion to be properly drawn?”

  • Type: Sufficient Assumption (SA)
  • Key insight: The right answer guarantees the conclusion by supplying a missing link (often a conditional bridge).

Common variation: “Which of the following, if assumed, enables the conclusion…?” is typically SA even without the word “sufficient.”

3) Strengthen vs Weaken (both often use “if true”)

Strengthen stem: “Which of the following, if true, most strengthens the argument?”

  • What to look for: a fact that supports the conclusion by:
    • bolstering a premise
    • connecting premise to conclusion (closing a gap)
    • ruling out an alternative explanation

Weaken stem: “Which of the following, if true, most seriously weakens the argument?”

  • What to look for: a fact that undermines by:
    • providing a counterexample
    • suggesting an alternative cause
    • attacking a key assumption

Fast ID tip: Strengthen/Weaken are “effect” questions: you judge answers by what they do to the argument.

4) Resolve / Explain a Discrepancy (Paradox)

Stem: “Which of the following, if true, most helps to resolve the apparent discrepancy?”

  • Type: Resolve
  • Stimulus pattern: two facts that seem inconsistent.
  • What the right answer often does:
    • introduces a distinction (different groups/conditions)
    • adds a time shift (before vs after)
    • clarifies a definition shift
    • identifies a hidden variable

Trap: Answers that attack one fact (weaken) rather than making both facts compatible.

Common Mistakes & Traps

  1. Misreading “EXCEPT/LEAST/NOT.”

    • What goes wrong: You pick the best strengthen/MBT answer.
    • Why it’s wrong: The question asks for the one that doesn’t.
    • Fix: Circle/underline the modifier and rephrase: “Find the odd one out.”
  2. Confusing MBT with MSS.

    • What goes wrong: You demand deductive certainty on MSS (or accept “pretty likely” on MBT).
    • Why it’s wrong: These have different proof standards.
    • Fix: For MBT, ask: “Is this forced?” For MSS, ask: “Is this the best-supported?”
  3. Treating Strengthen/Weaken like Inference.

    • What goes wrong: You reject helpful new information because it wasn’t in the stimulus.
    • Why it’s wrong: Strengthen/Weaken answers are new facts (introduced by “if true”).
    • Fix: Shift mindset: you’re testing impact, not deriving.
  4. Mixing up Necessary vs Sufficient Assumption.

    • What goes wrong: You choose a too-strong “guarantees it” answer on NA, or a too-weak “kind of helps” answer on SA.
    • Why it’s wrong: NA must be required; SA must be enough.
    • Fix: NA: try negation (does argument collapse?). SA: does it complete the proof?
  5. Ignoring what the question asks you to describe (Method/Flaw/Role).

    • What goes wrong: You evaluate whether the argument is persuasive instead of describing structure.
    • Why it’s wrong: These are descriptive tasks, not “fix the argument.”
    • Fix: Summarize in abstract terms: premises → (gap) → conclusion; then match wording.
  6. Parallel reasoning without matching logical form.

    • What goes wrong: You pick the answer with similar topic or vibe.
    • Why it’s wrong: Parallel questions are about structure (quantifiers, conditionality, causation).
    • Fix: Prephrase the skeleton (e.g., “All A are B; C is A; therefore C is B”).
  7. Point-at-issue questions: picking a statement both might agree/disagree with.

    • What goes wrong: You choose a nuanced statement neither clearly commits to.
    • Why it’s wrong: Correct answer must be a clean split.
    • Fix: For each option, force: Speaker 1 = yes/no; Speaker 2 = yes/no. You need opposite.
  8. Assuming every stimulus has a conclusion.

    • What goes wrong: You invent an argument and then answer the wrong task.
    • Why it’s wrong: Many stimuli are fact-sets.
    • Fix: If no claim is supported by others, treat it as inference (or discrepancy if it’s a puzzle).

Memory Aids & Quick Tricks

Trick / mnemonicWhat it helps you rememberWhen to use it
“EXCEPT = pick the ugly duckling”Four answers do the asked-for thing; one doesn’tAny stem with EXCEPT/NOT/LEAST/FALSE
Negation Test (NA)Necessary assumptions are required: negating them breaks the argumentNecessary Assumption questions
“MBT = Must Be Text-supported”Don’t add new facts; combine/restate only what’s givenMust Be True / Inference
“MSS = Most Supported, not Must”Softer standard; avoid demanding certaintyMost Strongly Supported
Conclusion hunt: ‘However/But’ = pivotThe real conclusion often follows contrast wordsAny argument stimulus
Parallel = form over contentIgnore subject matter; match logical skeletonParallel Reasoning / Parallel Flaw
Resolve = make both trueYou’re reconciling facts, not attacking themDiscrepancy/Paradox questions

Quick Review Checklist

  • You can identify whether the stimulus is an argument, a fact-set, or a discrepancy.
  • You scan stems for top signals: must, most strongly, if true, depends on, allows, flaw, method, parallel, principle, disagree, role.
  • You never miss EXCEPT/LEAST/NOT.
  • You apply the right proof standard:
    • MBT/NA = required
    • MSS = best-supported
    • Strengthen/Weaken/Resolve = impact-based
  • You distinguish NA vs SA by the task: required vs enough.
  • You treat Method/Flaw/Role as description problems, not “fix it” problems.
  • You remember: Parallel is structure, not topic.

You’ve got this—classification is the fastest way to stop LR from feeling random and start making every question predictable.