Logical Reasoning Question Types & Identification
What You Need to Know
Logical Reasoning (LR) is mostly pattern recognition: if you can correctly identify the question type, you immediately know (1) what the correct answer must do, (2) what wrong answers will look like, and (3) how to read the stimulus (argument vs. facts).
Core idea: every LR question has a “job”
- Argument-based questions: there is a conclusion supported by premises. Your job usually involves evaluating or repairing the reasoning.
- Fact-set questions: no argument is required. Your job is usually drawing a valid inference (or noticing what cannot be inferred).
The fastest classifier: 3 questions you ask yourself
- Is there a conclusion? (Look for a claim being supported.)
- What does the stem demand? (Strengthen? Must be true? Flaw?)
- Is the correct answer about truth or about reasoning?
- Truth: Must Be True / Most Strongly Supported / Cannot Be True.
- Reasoning: Strengthen/Weaken/Assumption/Flaw/Method/Parallel/Principle/Evaluate.
Critical reminder: You do not “do the problem” until you know the question type. Wrong type = wrong standard.
Step-by-Step Breakdown
This is the “night before” workflow to identify the type in seconds and aim your brain correctly.
1) Read the question stem first (label the type)
Use the stem to place the question into one of these buckets:
- Help the argument (Strengthen / Sufficient Assumption / Principle-Justify)
- Hurt the argument (Weaken)
- Find the missing link (Necessary Assumption)
- Diagnose the reasoning (Flaw / Method / Role / Main Point)
- Follow the facts (Must Be True / Most Strongly Supported / Cannot Be True)
- Resolve conflict (Resolve/Explain Discrepancy)
- Test the argument (Evaluate)
- Match structure (Parallel Reasoning / Parallel Flaw)
- State the dispute (Point at Issue)
2) Decide: argument or fact-set
- If the stem is inference (must be true / supported / cannot be true): treat the stimulus as a fact pattern even if it sounds persuasive.
- If the stem is strengthen/weaken/assumption/flaw: treat the stimulus as an argument even if it includes lots of facts.
3) For argument questions: isolate conclusion + support
Do a fast “core” markup in your head:
- Conclusion: what the author is trying to prove.
- Premises: reasons offered.
- Background: context, concessions, definitions.
Quick conclusion flags:
- therefore, thus, hence, so, consequently
- clearly, shows that, proves that
- recommendation/prediction (often the conclusion)
4) Pick the right “standard” for the question type
Different types demand different certainty:
- Must Be True: 100% forced by the facts.
- Most Strongly Supported: best supported (not necessarily forced).
- Strengthen/Weaken/Evaluate: “most helps/most hurts/most tests,” not “proves.”
- Necessary Assumption: must be true for the argument to work (but doesn’t have to guarantee the conclusion).
- Sufficient Assumption: if true, guarantees the conclusion.
5) Use the type-specific move
Here are the core moves by bucket:
- Strengthen: add a premise that supports the conclusion or closes a gap.
- Weaken: introduce an alternative cause, counterexample, or show premises don’t support conclusion.
- Necessary Assumption: use the Negation Test (negate answer choice; if argument collapses, it was necessary).
- Sufficient Assumption: look for a missing conditional/bridge that makes the conclusion follow.
- Flaw: describe the recurring error (causation, sampling, equivocation, etc.).
- Must Be True: combine statements conservatively; avoid adding assumptions.
Mini annotated example: Identify type → correct mindset
Stem: “Which one of the following, if true, most strengthens the argument?”
- Type: Strengthen (argument-based)
- Read stimulus for: conclusion + gap
- Correct answer: something that adds support for conclusion, often by:
- ruling out an alternative explanation
- showing premise is more reliable/representative
- linking premise to conclusion (new connecting fact)
Key Formulas, Rules & Facts
The high-yield question-type map (keywords → task)
| Question Type | Common stem language (tell-tales) | What the correct answer must do | Notes / traps |
|---|---|---|---|
| Must Be True (MBT) | “must be true,” “properly inferred,” “follows logically” | Be guaranteed by the statements | Don’t choose “likely” or “could.” No new info. |
| Most Strongly Supported (MSS) | “most strongly supported,” “best supported” | Be more supported than the rest | Can be probabilistic; still must be grounded in stimulus. |
| Cannot Be True | “cannot be true,” “must be false,” “least supported” (rarely) | Contradict the stimulus | Often a simple contradiction to one fact/constraint. |
| Main Conclusion / Main Point | “main conclusion,” “conclusion drawn,” “main point” | Identify the author’s conclusion | Beware intermediate conclusions. |
| Strengthen | “most strengthens,” “most supports,” “provides additional support” | Make conclusion more likely | Usually targets a gap/assumption. |
| Weaken | “most weakens,” “casts most doubt” | Make conclusion less likely | Alternative cause, counterexample, undermining a premise, showing mismatch. |
| Necessary Assumption | “assumption required,” “depends on,” “requires assuming” | Must be true for argument to work | Use Negation Test; answers often mild. |
| Sufficient Assumption (Justify) | “assumption that enables,” “if assumed, allows conclusion to be properly drawn” | If true, guarantees conclusion | Often strong; can be conditional bridge. |
| Flaw | “reasoning is flawed because,” “error in reasoning” | Correctly describe the logical mistake | Match to classic flaw families (see below). |
| Method of Reasoning | “argument proceeds by,” “method is to” | Describe how the argument reasons | Abstract description; keep it structural. |
| Role of a Statement | “the statement plays which role” | Identify function: premise, conclusion, objection, background | Find the referenced line; classify. |
| Parallel Reasoning | “most closely parallels,” “similar pattern of reasoning” | Same logical form (valid/invalid pattern) | Ignore topic; match structure + quantifiers/conditionals. |
| Parallel Flaw | “parallel in flawed reasoning” | Same type of mistake | First find flaw in stimulus, then match. |
| Principle (Strengthen) | “principle that supports/justifies” | A general rule that supports the argument | Often applied to case; must fit facts. |
| Principle (Justify) | “principle that, if valid, makes reasoning valid” | Principle that guarantees conclusion | Stronger than principle-strengthen. |
| Resolve / Explain Discrepancy | “helps explain,” “resolve the apparent conflict” | Make both facts compatible | Not strengthen/weaken; just reconcile. |
| Evaluate | “would be most useful to know,” “helps evaluate” | Identify info that would matter | Usually a yes/no question whose answers swing strength. |
| Point at Issue | “disagree about,” “point at issue” | Statement one would say yes, other no | Must be something both address. |
Classic flaw families (high-frequency)
| Flaw family | What it looks like | Fast diagnostic |
|---|---|---|
| Causation | Treats correlation/sequence as cause | “Is there an alternative cause? Reverse causation? Coincidence?” |
| Sampling/representativeness | Generalizes from biased/too small sample | “Is the sample typical of the population?” |
| Equivocation | Shifts meaning of a key term | “Same word used in different sense?” |
| Part–whole / composition–division | Attributes group trait to individuals or vice versa | “Does property transfer?” |
| Conditional confusion | Mistakes sufficient/necessary; affirms consequent/denies antecedent | “Did they reverse the arrow?” |
| Comparison/analogy | Assumes because similar in some ways, same in relevant way | “Are relevant differences ignored?” |
| Prescriptive (is–ought) | Moves from facts to recommendation without value/goal bridge | “Where’s the value premise?” |
The Negation Test (Necessary Assumption)
- Take a candidate assumption and negate it (make it as close to exact opposite as possible).
- If the negation makes the argument fall apart, the original was necessary.
- If the negation leaves the argument mostly intact, it wasn’t necessary.
Negation tips:
- “Some” ↔ “None”
- “All” ↔ “Not all”
- “Many” ↔ “Not many” (or “few,” context-dependent)
- “Usually” ↔ “Not usually”
Don’t over-negate: “All A are B” negates to “Not all A are B” (i.e., at least one A is not B), not “No A are B.”
Examples & Applications
Below are short “spot the type → do the right thing” drills.
Example 1: Necessary vs. Sufficient Assumption (stem tells you which)
Stimulus: “All of the committee members who reviewed the proposal supported it. Therefore, the committee will approve the proposal.”
If stem says “requires assuming” → Necessary Assumption.
- Gap: approval by the whole committee vs. support by reviewers.
- A necessary assumption could be: “Only committee members who reviewed the proposal will vote on it.” (or “All committee members reviewed it.”)
- Use Negation Test: If some non-reviewing members vote and oppose, conclusion can fail.
If stem says “if assumed, enables the conclusion” → Sufficient Assumption.
- You’d want a stronger bridge like: “The committee approves any proposal that all reviewing members support.”
Key insight: Same stimulus, different answer standards.
Example 2: Strengthen (rule out an alternative)
Stimulus: “After the city installed LED streetlights, nighttime accidents decreased. So the LED lights caused the decrease.”
Stem: “Which option, if true, most strengthens?”
Best strengthen patterns:
- Rule out alternatives: “No other traffic-safety changes occurred during that period.”
- Improve causal link: “Accident rates decreased most on streets that received LED lights earliest.”
Wrong answer smell:
- Restates the conclusion (“LED lights are good”) without connecting evidence.
Example 3: Resolve/Explain Discrepancy (not strengthen/weaken)
Stimulus: “A medicine performed better than placebo in early trials. But in a larger trial, it performed no better than placebo.”
Stem: “Which option best resolves the apparent discrepancy?”
Correct answer must make both true, e.g.:
- Early trials used patients with severe symptoms; larger trial included many mild cases where improvement is hard to detect.
Trap:
- An answer that says “the medicine doesn’t work” is not a resolution; it picks a side rather than reconciling.
Example 4: Point at Issue (force a yes/no split)
Stimulus: Speaker A: “The company should allow remote work because it increases productivity.”
Speaker B: “Remote work should not be allowed because it harms collaboration.”
Stem: “They disagree about which issue?”
A correct “point at issue” statement would be something like:
- “Allowing remote work increases overall productivity.”
- A: yes. B: not necessarily/likely no.
Trap:
- A statement like “Collaboration is valuable” may be something both agree with, so it’s not the dispute.
Common Mistakes & Traps
Mistake: Treating MSS like MBT
- What you do: demand 100% proof for “most strongly supported.”
- Why wrong: MSS is a comparative standard; the best-supported answer can be less than certain.
- Fix: ask “Which choice is most anchored in the facts?” not “Which must be true?”
Mistake: Confusing Necessary vs. Sufficient Assumption
- What you do: pick a strong “guarantee” statement for a necessary-assumption stem (or vice versa).
- Why wrong: Necessary assumptions are often minimal; sufficient assumptions can be powerful bridges.
- Fix: use the Negation Test only for necessary; for sufficient, ask “If this is true, does the conclusion have to follow?”
Mistake: Strengthen/Weaken by repeating a premise
- What you do: choose an answer that rephrases something already stated.
- Why wrong: it rarely changes probability; it’s usually redundant.
- Fix: look for answers that address the gap (new link, alternative explanation, representativeness).
Mistake: Doing Flaw questions like Weaken questions
- What you do: pick an answer that would weaken the argument if true.
- Why wrong: Flaw asks for a description of the error, not new information.
- Fix: translate the argument into a simple skeleton and name the mistake (causal, sampling, equivocation, conditional, etc.).
Mistake: Missing the conclusion (especially with concessions)
- What you do: treat “although…” or “some people say…” as the author’s view.
- Why wrong: those are often opposing views or background.
- Fix: ask “What is the author trying to get you to believe by the end?”
Mistake: Over-assuming in inference questions
- What you do: bring in real-world knowledge or “common sense” beyond the text.
- Why wrong: MBT/Cannot be True require strict textual support.
- Fix: be literal; combine facts conservatively.
Mistake: Parallel questions matched by topic, not structure
- What you do: pick the answer with the most similar subject matter.
- Why wrong: parallel is about form, not content.
- Fix: strip nouns → variables; match quantifiers/conditionals and whether reasoning is valid/invalid.
Mistake: Resolve questions answered like Strengthen/Weaken
- What you do: pick an answer that supports one side.
- Why wrong: resolve must make both statements coexist.
- Fix: ask “Could both facts still be true if this answer is true?”
Memory Aids & Quick Tricks
| Trick / mnemonic | What it helps you remember | When to use it |
|---|---|---|
| “Stem = Steering Wheel” | The stem controls your standard and approach | Always read stem first |
| Negate to Verify (NA) | Necessary assumptions are exposed by negation | Necessary Assumption questions |
| “Strengthen = Shore up a gap” | Most strengthen answers address a missing link or alternative | Strengthen / Principle-support |
| “Weaken = Another reason / Not representative / Not causal” | Three common weaken routes | Weaken questions |
| “Resolve = Make both true” | You’re reconciling, not proving | Discrepancy/Paradox questions |
| “Parallel = Variables, not vibes” | Replace content with letters to match form | Parallel Reasoning/Flaw |
| “Evaluate = Ask what would swing it” | Correct answer usually identifies a key unknown with two possible outcomes | Evaluate questions |
| “Point at issue = Yes/No split” | One speaker agrees, the other disagrees | Point at Issue questions |
Quick Review Checklist
- You can label the question type from the stem before reading deeply.
- You know whether you’re in truth mode (MBT/MSS/Cannot) or reasoning mode (strengthen/weaken/assumption/flaw/etc.).
- For argument questions, you can identify conclusion vs. premises fast.
- You apply the correct certainty standard:
- MBT = forced
- MSS = best supported
- Strengthen/Weaken/Evaluate = most affects
- NA = required (Negation Test)
- SA = guarantees
- You recognize the big flaw families: causal, sampling, equivocation, conditional, analogy, part–whole, is–ought.
- You don’t confuse Resolve with Strengthen/Weaken: resolve makes both facts compatible.
- You treat Parallel as structure-matching, not topic-matching.
You’ve got this—if you nail identification, the right answers start looking inevitable.