Inference & Application Questions in RC
1. What You Need to Know
Inference & Application questions are where LSAT Reading Comp rewards proof-based reading, not “what sounds likely.” They ask you to (a) draw a conclusion that is supported by the passage or (b) apply the passage’s principle/framework to a new scenario.
What counts as an inference in RC?
An RC inference is a claim that is logically supported by what the passage says (often across multiple sentences), even if the passage doesn’t say it verbatim.
- Think: Must be true / most strongly supported given the passage.
- Your job: find textual support and avoid adding new assumptions.
What counts as an application in RC?
An application question gives you a new situation and asks which option best matches or follows the passage’s reasoning, rule, standard, or distinction.
- Think: Use the passage like a rulebook.
- Your job: identify the relevant generalization/criterion in the passage, then test choices for fit.
Bottom line: These questions are won by (1) anchoring yourself to specific lines/ideas in the passage and (2) policing “strength” (too strong = usually wrong).
Common stems you’re looking for
Inference-type stems
- “The passage most strongly supports which of the following?”
- “It can be inferred that…”
- “The author would be most likely to agree that…”
- “Which of the following is most strongly suggested by the passage?”
- “Which statement is most supported by the passage?”
Application-type stems
- “Which of the following is most analogous to…?”
- “Which scenario best illustrates the principle described?”
- “Which of the following best conforms to the approach/criterion/standard?”
- “Based on the passage, how would the author most likely evaluate…?”
2. Step-by-Step Breakdown
A. Inference Questions (Must Be True / Most Supported)
Label the task: prove, don’t predict.
- Tell yourself: “I need a choice with a receipts trail in the passage.”
Locate the anchor.
- Many inference questions point to a paragraph, concept, or viewpoint. If not, use keywords from answers to decide where to look.
Prephrase a low-resolution inference.
- Aim for a safe, modest takeaway (often a paraphrase-level conclusion).
- Good prephrase style: “Seems like X is generally true,” “The author likely thinks Y is limited,” “Viewpoint A would reject B because…”
Test answers with a “support standard.”
- For each choice, ask: Where is this supported?
- If you can’t point to a line/idea, it’s out (even if it sounds reasonable).
Police strength and scope.
- Eliminate answers that are:
- Stronger than the passage (all/never/proves/guarantees)
- Broader than the passage (extends to new domains)
- More certain than the passage (turns “may” into “will”)
- Eliminate answers that are:
Cross-check against the passage’s structure.
- Correct inferences often track:
- author’s main point
- a concession (“although…”) and its limit
- the passage’s comparisons/distinctions
- a viewpoint’s reasoning pattern
- Correct inferences often track:
Micro-example (inference workflow)
If the passage says: “Some researchers argue X improves outcomes; however, multiple studies find the effect disappears when controlling for income.”
- Safe inference: “The effect of X may depend on income (or may not be robust once income is controlled).”
- Trap inference: “X does not improve outcomes.” (too strong)
B. Application Questions (Principle / Analogy / New Scenario)
Identify the passage rule/standard.
- Ask: “What is the criterion the passage uses to classify/judge something?”
- Often signaled by language like “distinguishes,” “criterion,” “depends on,” “only if,” “generally,” “tends to,” “is considered”.
Extract the ‘moving parts’ (the checklist).
- Turn the rule into 2–4 concrete features.
- Example format: “For something to count as \text{Category A}, it must have Feature 1 and Feature 2, and it must not have Feature 3.”
Map the new scenario onto the checklist.
- Don’t be hypnotized by surface similarity (same topic words).
- Ask: “Does this scenario satisfy the same functional role in the argument?”
Eliminate answers that miss a required element or violate a restriction.
- Application questions are often won by one missing condition.
Pick the best fit, not a perfect clone.
- If none are perfect, choose the option that matches the most important features (the ones the passage emphasizes as decisive).
Micro-example (application workflow)
If the passage’s principle is: “A regulation is effective only if compliance is easy to verify.”
- In a new scenario, the best match will involve a rule where verification feasibility is the key determinant—not whether the rule is popular.
Critical reminder: Application questions are still text-bound. You’re applying the passage’s framework, not your personal beliefs.
3. Key Formulas, Rules & Facts
Inference vs. Application: what you’re proving
| Question type | What you’re trying to do | What “wins” | Common wrong-answer vibe |
|---|---|---|---|
| Inference (Most Supported / Must Be True) | Derive a claim supported by the passage | Conservative statement with clear support | Sounds plausible but not provable; too strong; outside scope |
| Application (Principle / Analogy / Evaluate scenario) | Use passage’s criterion to classify/evaluate a new case | Best fit with the passage’s decisive features | Superficial resemblance; misses a condition; swaps roles |
Support ladder (how far you can go)
- Direct text (verbatim or near-verbatim) = safest
- Paraphrase (same meaning, different words) = safe
- Synthesis (combines two supported ideas) = safe if each piece is supported
- Interpretive inference (explains an implication) = okay only if it’s forced
- Speculation (requires new assumptions) = wrong
Strength & scope rules (high-yield)
| Signal in answer choices | Typical issue | What you do |
|---|---|---|
| All, none, never, always, proves, guarantees | Too strong for RC support | Eliminate unless passage is equally absolute |
| Some, often, can, may, tends to | More matchable to passage’s cautious claims | Prefer when passage is nuanced |
| Broad categories (“all scientists,” “the public,” “all markets”) | Scope creep | Check if passage actually generalizes that far |
| New concept not discussed | Outside scope | Eliminate unless it’s a necessary bridge |
Where correct inferences often come from (passage “hot spots”)
- Concessions: “Although X, Y…” → inference about limits of X or priority of Y
- Comparisons: “Unlike A, B…” → inference about distinguishing features
- Causal cautions: “correlated with,” “may contribute,” “not necessarily” → inference that causation is uncertain/conditional
- Author attitude markers: “surprisingly,” “importantly,” “mistakenly,” “overlooks” → inference about evaluation
- Resolution of a debate: which side is weakened/qualified and why
Application mapping rules (what must match)
| Passage element | What to match in the answer choice | Watch-outs |
|---|---|---|
| Criterion/definition | Required features and prohibitions | Missing one required feature = out |
| Role in argument | What the example is doing (supporting, criticizing, explaining) | “Same topic” but different argumentative role |
| Tradeoff/priority | Which value wins when they conflict | Answers that flip the priority |
| Conditional logic (if/only if) | Direction of requirement | Mixing up necessary vs sufficient |
You don’t need formal logic symbols to win RC, but you must respect directionality: “Only if verified” means verification is required, not that it guarantees effectiveness.
4. Examples & Applications
Example 1: Straight inference (author’s view)
Passage snippet: “While early critics dismissed the method as impractical, later implementations reduced costs dramatically, though they still require specialized training.”
Most supported inference: The method is less costly than it used to be, but it still isn’t universally accessible due to training requirements.
Key insight: Combine two supported points (costs down + training still needed) without overreaching (don’t infer it’s now “widely adopted” unless stated).
Example 2: Inference from a concession
Passage snippet: “Some historians argue the policy stabilized the region. However, the passage notes that stability coincided with external economic growth and may not have resulted from the policy itself.”
Most supported inference: The passage suggests the policy’s role in causing stability is uncertain.
Variation it tests: Students often jump from skepticism to reversal (“policy did not stabilize”). The passage only casts doubt; it doesn’t disprove.
Example 3: Application of a criterion
Passage principle: “The passage distinguishes ‘restoration’ from ‘reconstruction’: restoration aims to return an artifact to a prior state using period-accurate materials; reconstruction permits modern substitutes if they preserve appearance.”
Question: Which scenario is an example of restoration?
Correct setup: Look for (1) prior state + (2) period-accurate materials. If the scenario uses modern substitutes—even if it looks identical—it’s reconstruction.
Key insight: Don’t be distracted by “looks the same.” The decisive feature is materials standard.
Example 4: Analogy / parallel reasoning in RC
Passage reasoning pattern: “Because the metric is easily gamed, performance measured by that metric will not reliably indicate true improvement.”
Best analogous scenario: A system where a measurable proxy can be manipulated, making the measured gains unreliable (e.g., teaching to a narrow test that boosts scores without improving underlying skill).
Key insight: Match the structure: proxy measure \rightarrow incentive to game \rightarrow unreliable indicator. Not just “something about measurement.”
5. Common Mistakes & Traps
Mistake: Treating inference as ‘most likely true in the real world.’
- What goes wrong: You pick the most believable statement.
- Why wrong: LSAT wants what is supported by the text, not general knowledge.
- Fix: Demand a line/idea citation for every claim.
Mistake: Overstrengthening (making the passage more confident than it is).
- What goes wrong: Turning “may,” “can,” “suggests” into “will,” “proves,” “always.”
- Why wrong: RC authors hedge; correct answers mirror that.
- Fix: Match the passage’s certainty level.
Mistake: Scope creep (exporting a limited point to a broad conclusion).
- What goes wrong: Passage discusses a specific context; answer generalizes to all contexts.
- Why wrong: Inference must stay within the passage’s defined universe.
- Fix: Ask: “Did the passage really talk about all of that?”
Mistake: Confusing a viewpoint’s claim with the author’s endorsement.
- What goes wrong: Attributing a position to the author when it’s just reported.
- Why wrong: RC often presents multiple viewpoints neutrally or critically.
- Fix: Track author stance using evaluative words (e.g., “fails,” “overlooks,” “compelling”).
Mistake: For application, matching topic words instead of the passage’s criterion.
- What goes wrong: You choose the answer that “sounds like the passage subject.”
- Why wrong: Application is about features/structure, not vocabulary.
- Fix: Build a checklist from the passage and test each choice against it.
Mistake: Missing a single required condition in application questions.
- What goes wrong: An answer matches 2 features but violates the decisive constraint.
- Why wrong: LSAT designs wrong answers to be “almost right.”
- Fix: Identify which condition is necessary (often signaled by “only if,” “must,” “depends on”).
Mistake: Reversing relationships (cause/effect, problem/solution, criterion/outcome).
- What goes wrong: You select an option that flips what leads to what.
- Why wrong: Many RC inferences hinge on directional claims.
- Fix: Re-state the relationship in simple terms before evaluating choices.
Mistake: Treating an example as the rule (or treating the rule as only that example).
- What goes wrong: You either overgeneralize from one example or refuse to generalize at all.
- Why wrong: RC often uses examples to illustrate a broader claim.
- Fix: Ask: “Was the example presented as illustrative or as exhaustive?”
If you’re torn between two answers, the winner is usually the one with less extra content beyond what the passage supports.
6. Memory Aids & Quick Tricks
| Trick / mnemonic | What it helps you remember | When to use it |
|---|---|---|
| “Prove it, don’t improve it.” | Don’t strengthen/embellish the passage | Any inference question |
| Low-res prephrase | Start with a modest takeaway so you don’t overreach | Before looking at answer choices |
| “Receipts test” | If you can’t point to support, it’s out | When an answer sounds tempting |
| Checklist extraction | Turn a principle/definition into 2–4 required features | Application / principle questions |
| Role-match | Match the function (proxy, incentive, limitation, exception) rather than the topic | Analogy / parallel application |
| Strength dial | Compare passage vs answer on certainty (may vs must) | When choices differ by intensity |
| “Same universe?” | Keep scope consistent (time period, group, context) | Any inference/application choice that feels broad |
7. Quick Review Checklist
- Inference = must be supported (often the most conservative supported choice).
- For any answer you like, ask: Where is the support? If you can’t locate it, drop it.
- Match strength: hedge-to-hedge, absolute-to-absolute.
- Match scope: don’t expand groups, timeframes, or domains.
- Separate author view from others’ views.
- Application = passage rulebook: extract the criterion, make a checklist, test each choice.
- In analogy/application, match structure/role, not just shared buzzwords.
- When stuck: prefer the option that says less and stays closer to the passage.
You’ve got this: stay text-anchored, keep your inferences modest, and let the passage do the work.