LSAT

1. Find the conclusion before judging it

Learning goal: Distinguish evidence from the claim it supports.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

Logical Reasoning asks you to analyze arguments, including premises, conclusions and assumptions [L2]. A conclusion is the claim the speaker wants you to accept because of other claims. It need not be the last sentence. A test is to ask whether "because" can introduce one statement as support for another. Background and concessions can matter without being the central conclusion.

Consider a city arguing that a bus route should be extended because evening ridership is rising. The ridership claim is evidence; extension is a recommendation. Whether extension is a good policy is a separate issue. Identify the structure before importing your own transport views, then ask whether the stated evidence actually bridges to the recommended action.

Worked example

"The museum should stay open later. Evening requests have doubled, and the current closing time excludes commuters." Identify conclusion and support.

1. The proposed action is later opening.
2. Doubled requests and commuter exclusion are offered as reasons.
3. Therefore the first sentence is the main conclusion.
4. Staffing feasibility is a possible gap, not a stated premise.

Mental model: Separate the claim from its support before evaluating the gap.

Common trap: An unstated assumption is not a quoted premise.

2. Arrow direction matters

Learning goal: Use contraposition without reversing an implication.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

"If A then B" says A is sufficient for B and B is necessary for A. It does not say B is sufficient for A. Its contrapositive, if not B then not A, preserves the relationship. The converse B → A and inverse not A → not B do not follow without more information. LSAC includes conditional reasoning within LR [L2].

Use a counterexample rather than confidence in a familiar sentence. If all archivists have training, a trained non-archivist is possible. That person disproves the converse while preserving the premise. Universal statements do not assert that an archivist exists; this distinction becomes important when an answer adds "at least one."

Worked example

All archivists have training. No trained person is careless. What must follow?

1. Archivist → trained.
2. Trained → not careless.
3. Chain the implications: archivist → not careless.
4. No archivist is careless. Existence and the reversed training implication remain unproved.

Mental model: Chain valid arrows; contrapose by reversing and negating.

Common trap: Never reverse an arrow because the words seem related.

3. Some needs a witness; all needs every case

Learning goal: Preserve the exact quantifier in an inference.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

Every funded pilot has oversight, and some pilots are funded. The funded pilot supplies a witness that is both a pilot and overseen. Therefore some pilots have oversight. You cannot conclude that all pilots do: an unfunded pilot without oversight can coexist with the premises. This is an original LR inference exercise, not a resurrected logic-games section [L2].

"Some" means at least one, not necessarily a minority; "most" means more than half in the specified group. Two "some" statements do not guarantee overlapping witnesses. If some artists are runners and some runners are chefs, the artist-runner and chef-runner can be different people. Keep the universe and reference group fixed.

Worked example

Every funded pilot has oversight. Some pilots are funded. Does every pilot have oversight?

1. Let pilot A be funded; A must have oversight.
2. Add pilot B who is unfunded and has no oversight.
3. Both premises remain true.
4. All pilots overseen is false in this countermodel; some pilots overseen remains true.

Mental model: A witness inherits properties; universals do not create witnesses.

Common trap: Do not upgrade some to all.

4. Necessary is not enough

Learning goal: Test an assumption by negating it precisely.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

A necessary assumption must hold for the argument to work as offered. Negating it should undermine that reasoning. A sufficient assumption, by contrast, is strong enough to make the conclusion follow; it need not be indispensable. LR tests assumptions and evaluation [L2]. Read the stem before choosing the strongest-sounding bridge.

An argument that recommends a new night bus because commuters cannot currently get home needs a link between the proposed bus and the commuters' problem. It does not necessarily need every commuter to ride. Negate quantifiers carefully: "all" negates to "not all," and "some" negates to "none." Replacing "some" with "few" is not a genuine negation.

Worked example

"We should create a night bus to help stranded commuters; it will serve their journey." Test: at least some stranded commuters can use the route.

1. The recommendation relies on a benefit to stranded commuters.
2. Negate the proposed assumption: none can use the route.
3. Then the claimed route benefit fails.
4. This is a required bridge, unlike the stronger claim that every commuter must use it.

Mental model: Necessary assumptions survive neither precise negation nor scope inflation.

Common trap: A helpful bridge is not automatically a required bridge.

5. Give the cause a rival

Learning goal: Explain an association without the proposed cause.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

A causal argument can confuse correlation with cause, reverse the direction, or overlook a common cause. LSAT LR asks how evidence strengthens or weakens arguments [L2]. A useful weakening answer describes how the reported association could arise even if the proposed cause were absent or smaller than claimed. An irrelevant criticism is not enough.

If workplaces with wellness classes have fewer absences, selection is a rival explanation: healthier or more highly resourced workplaces might be more likely to offer classes. To strengthen causation, a matched or randomized comparison can address that alternative, subject to the facts supplied. Do not treat any study label as a guarantee of correctness.

Worked example

Cities with more parks have healthier residents; therefore parks explain the entire health gap. Suggest a weakening finding.

1. The observed facts are park availability and health.
2. The claim gives the entire gap to parks.
3. If income is higher in park-rich cities and affects health, it is a possible common cause.
4. Evidence of that pre-existing group difference weakens the exclusive causal explanation.

Mental model: Explain how the rival produces the observation.

Common trap: A mere possibility matters only if it connects to the argument.

6. Answer the stem, not a true statement

Learning goal: Choose evidence that does the requested logical work.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

LSAC explicitly advises not selecting an answer merely because it is true and using the information supplied [L6]. A strengthening stem asks what improves support, an inference stem asks what follows, and a paradox stem asks what reconciles observations. The same fact can be relevant to one task and useless to another.

Suppose a library reports more digital loans but fewer building visits. A reconciliation answer might say more patrons access material remotely. That need not prove the library is financially healthy, which is a different conclusion. Before reading choices, describe the job an answer must do in one sentence, then test each option against it.

Worked example

Paradox: library loans rose while visits fell. Which fact could reconcile both?

1. Accept both reported changes rather than disputing them.
2. Loans can occur digitally without a building visit.
3. If digital access expanded, loans can rise as visits fall.
4. This supplies a joint explanation, not a claim about profit or reading quality.

Mental model: Truth, relevance and task fit are separate checks.

Common trap: Do not solve an easier question than the one asked.

7. Map the passage by function

Learning goal: Track how paragraphs build the author's position.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

Reading Comprehension tests structure, purpose, tone and relationships, not only factual recall [L3]. Mark each paragraph's job: introducing a dispute, presenting an old view, challenging it, giving evidence, or qualifying a new position. This map helps locate detail questions and prevents a quoted opponent from becoming the author's thesis.

The current RC description permits variable single/comparative-set configurations [L3]; do not assume exactly one paired set on every form. These mini-passages isolate structure, not official passage length. The aim is to link the main claim to its support while preserving reservations such as incomplete evidence or limited scope.

Worked example

P1: an old theory attributes urban growth only to trade. P2: transport evidence supports another cause. P3: records remain incomplete. What is P3 doing?

1. P1 states the view under examination.
2. P2 broadens the proposed explanation.
3. P3 flags a limit in the evidence.
4. P3 qualifies confidence; it need not restore the trade-only theory.

Mental model: Map paragraph jobs, not only topics.

Common trap: The author can discuss a view without endorsing it.

8. Compare passages on the same question

Learning goal: Find agreement and disagreement with precise scope.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

When a paired set appears, compare what both texts actually address. Two writers may agree that a method has benefits but disagree about whether its costs justify broad adoption. Choosing "total disagreement" discards the shared premise. RC relationship questions require exact comprehension of both passages [L3].

Make a small relation map: shared topic, A's claim, B's claim, and evidence each would accept. A passage that never discusses cost cannot be assigned a cost position simply because its tone is enthusiastic. The original cases below deliberately vary the relation so that switching cases changes the warranted conclusion.

Worked example

A says online archives improve access but require labels. B says online archives improve access and labels deserve funding. Where do they agree?

1. Both explicitly endorse improved access.
2. Both treat labeling as important to use.
3. Neither supplied text claims digitization is sufficient on its own.
4. Their shared position is access benefit plus the importance of labeling.

Mental model: Compare positions on a shared question with matched scope.

Common trap: Silence and weaker claims do not establish contradiction.

9. Infer with the smallest justified step

Learning goal: Use countermodels to reject overstrong answers.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

An inference must stay within the passage's evidence. If a report says a policy sometimes reduces delay, it does not follow that it always reduces delay or that it is the cheapest policy. LR and RC both reward control of scope [L2, L3]. The best answer is not necessarily the most ambitious or the one that matches your outside knowledge.

To test necessity, imagine a world where all supplied statements hold but the candidate answer fails. If such a countermodel is possible, the answer is not required. You need not enumerate every world to refute one overstatement: a single legitimate exception defeats an all claim, and a second unranked policy defeats a cheapest claim.

Worked example

All audited firms keep records. Firm K keeps records. Must K be audited?

1. Premise: audited → records.
2. K satisfies records, the necessary condition.
3. A non-audited firm may also keep records.
4. K could be such a firm, so auditing does not follow.

Mental model: Refute a must-follow answer with one valid countermodel.

Common trap: A reasonable inference still needs support in the supplied text.

10. Writing a position that survives an objection

Learning goal: Defend a thesis and address a supplied perspective.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

Argumentative Writing is separately remotely proctored and unscored; an approved sample is required before score release. Current specifications provide 15 minutes for prewriting analysis, with optional advance after five, then 35 minutes to write. The task requires a position and engagement with at least one supplied perspective [L1]. This is not a legacy logic-games or scored essay section.

A strong practice outline connects thesis, reason and example, then answers a real objection. Merely saying "critics are wrong" does not explain why your tradeoff is justified. Avoid invented factual evidence: use an explicitly hypothetical example or the prompt's material. For multiple-choice pacing, there are four 35-minute sections and a ten-minute intermission after section two [L1], not an invented fixed question total.

Worked example

Practice issue: fund longer library hours or more digital access. Outline a position favouring digital access while addressing users without reliable internet.

1. Thesis: prioritize digital access while preserving a staffed access option.
2. Reason: remote access can serve users outside opening hours.
3. Objection: internet gaps exclude some users.
4. Response: reserve part of the plan for library terminals or lending access devices; acknowledge cost rather than claiming a free solution.

Mental model: Thesis → support → fair objection → relevant response.

Common trap: Unscored writing still has a score-release requirement.

11. Negating a requirement tests necessity

Learning goal: Test whether removing a proposed assumption damages the conclusion.

An assumption can be necessary without being enough to prove a conclusion. To test a proposed necessary assumption, negate it carefully and ask whether the argument can still function. Do not replace a modest statement with an extreme opposite. The negation of some is none; the negation of all is at least one not. Scope matters as much as vocabulary.

Original argument: a committee says its survey establishes resident support for a park because eighty percent of respondents favor it. A proposed assumption says respondents are not so unrepresentative that their preference tells us nothing about residents. Negating that assumption undermines the move from respondents to the resident population. By contrast, assuming every resident favors the park is far stronger than needed.

The negation test is a diagnostic method, not an automatic proof for every option. First identify the conclusion and its bridge. A premise about observed survey answers supports a broader population claim only with some reason to treat those answers as informative. A statement about the color of the survey paper does not repair that bridge unless the argument links it to response behavior.

Practice set: the museum expects attendance to rise because every visitor who was asked preferred weekend opening. Which bridge is more relevant: the interviewed visitors provide some evidence about potential attendance, or every citizen likes museums? The first is closer to the inferential need. Now negate it: the interviews provide no attendance evidence. Explain why the stated support weakens. The example is invented and is not an official LSAT item.

Source alignment: [L2/L3]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

Respondents prefer a park; conclusion says residents support it. Test the representativeness bridge.

1. Separate respondents from residents.
2. Locate the generalization.
3. Negate the claim that responses are informative.
4. The support for the population claim is damaged.

Mental model: Test whether removing a proposed assumption damages the conclusion.

Common trap: Negate a quantified claim with an unnecessarily extreme opposite.

12. A principle needs the right direction

Learning goal: Apply a stated sufficient condition without reversing it.

A principle can state that satisfying a condition is enough to justify a result. That does not mean the result proves the condition occurred. Translate the principle into a conditional before applying it. If a proposal lowers cost without reducing access, approve it means the two conditions together are sufficient for approval. Approval alone does not establish either condition.

In an original case, proposal A lowers cost and preserves access. The principle supports approving A. Proposal B preserves access but has unknown cost. The principle alone does not settle B. Proposal C has been approved, but nothing says it lowers cost. Other grounds for approval may exist. These distinctions prevent a matching option from slipping in a reversed arrow.

For a principle-justify task, the question may ask which rule connects the premises to the conclusion. For an application task, the principle is supplied and you must find a case that fits it. Read the stem before choosing. A broader principle can sometimes justify a conclusion, but an option that contradicts a stated condition cannot be rescued by being generally sensible.

Practice set: principle says records that are both incomplete and unverified should not be accepted. An incomplete but verified record does not satisfy both conditions. A complete but unverified record also does not. An incomplete and unverified record does. Explain why and matters: replacing it with or changes the rule. These cases practice reasoning from a supplied principle rather than giving advice about real institutional policies.

Source alignment: [L2/L3]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

Rule: if cost falls and access is preserved, approve. A satisfies both.

1. Treat both conditions as a conjunction.
2. A lowers cost.
3. A preserves access.
4. The rule supports approval.

Mental model: Apply a stated sufficient condition without reversing it.

Common trap: Reverse the principle or change and to or.

13. Resolve a puzzle without denying the facts

Learning goal: Find a distinction that lets both observations remain true.

A resolve-the-discrepancy question gives observations that seem difficult to reconcile. A useful answer explains how both can be true. It does not merely reject one of them or restate the surprise. List the two observations and name the assumption that makes them seem incompatible. The resolution often changes a denominator, a comparison group or a mechanism.

Original puzzle: a shop lowered its price, yet its total revenue increased. There is no contradiction if unit sales grew enough. Revenue equals price times quantity. A fall from ten to eight per item paired with growth from one hundred to two hundred items increases total revenue from one thousand to sixteen hundred. The example shows a possible explanation, not proof that sales actually changed in an unspecified real shop.

An option saying revenue did not increase contradicts the given fact rather than explaining it. An option saying customers care about prices may be related but does not state the needed quantity change. Prefer an answer that directly supplies the missing relationship. If the puzzle concerns a rate and a count, inspect the denominator instead of assuming they must move together.

Practice set: a town reports more cycle crashes but a lower crash rate per cyclist. Both can hold if the cyclist population rises enough. In invented numbers, ten crashes among one thousand cyclists and fifteen among three thousand yield rates of one percent and one-half percent. A resolution does not prove the policy is safe, causal or successful on every metric. It only removes the apparent conflict. State the limit of the explanation.

Source alignment: [L2/L3]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

Price fell 10 to 8; quantity rose 100 to 200. Explain higher revenue.

1. Keep both facts.
2. Revenue is price times units sold.
3. Earlier revenue 1000; later 1600.
4. Quantity growth outweighs price decline.

Mental model: Find a distinction that lets both observations remain true.

Common trap: Resolve a discrepancy by denying a stated fact.

14. Parallel flaws depend on structure

Learning goal: Match the reasoning error rather than the subject matter.

A parallel-reasoning question may use unfamiliar topics, but its logical structure can be short. Strip away the nouns and write the relation. If every A is B and an object is B, concluding it is A reverses a conditional. An option about the same subject may have valid reasoning and be a poor match. An option about a different subject may repeat the same invalid structure.

Original flawed argument: every approved application has a reference; this application has a reference; therefore it is approved. The structure is approval implies reference, reference, therefore approval. Having a reference is necessary under the premise but not enough for approval. The flaw is affirming the consequent. Do not confuse it with denying the antecedent, where absence of a sufficient condition is taken as absence of its result.

Compare: every oak is a tree; this is a tree; therefore it is an oak. That matches the flaw. Every oak is a tree; this is an oak; therefore it is a tree is valid and does not match. Every oak is a tree; this is not a tree; therefore it is not an oak is a valid contrapositive. Diagramming keeps the distinctions visible.

Practice set: every winner received a certificate; Jo received a certificate; therefore Jo won. Find the matching pattern in a new topic and explain why the certificate alone is not enough. Then change the final evidence to Jo did not receive a certificate: the premise now supports Jo did not win. Same nouns, different structure. The invented examples practice conditional analysis, not official test content.

Source alignment: [L2/L3]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

Approved implies reference. Reference observed. Approval concluded.

1. Translate approval → reference.
2. The evidence supplies the consequent.
3. The conclusion reverses the arrow.
4. The consequent can arise without approval.

Mental model: Match the reasoning error rather than the subject matter.

Common trap: Match topical nouns instead of logical structure.

15. A passage viewpoint needs attribution

Learning goal: Separate author judgments from positions the author reports.

A reading passage can contain several voices. Mark where the author states a view, where a scholar or critic is quoted or paraphrased, and where the author responds. A reported claim is not automatically endorsed. Verbs such as argues, questions and concedes signal relationships between ideas, but the surrounding explanation determines the author attitude.

Original passage: some critics argue that public archives are obsolete because digital search is faster. The author accepts that search tools reduce retrieval time but says archives also preserve originals and provenance. The author therefore favors combining search access with archival preservation. The critic conclusion and the author conclusion differ despite sharing a premise about speed.

A correct main-point answer should reflect the whole passage, not a vivid detail. Faster retrieval is part of the discussion, not the final claim that archives should disappear. An attitude answer claiming the author rejects every digital tool is too strong because the author concedes their benefit. Scope and concession help calibrate the position.

Practice set: classify three statements as critic, author or shared observation. Archives are obsolete belongs to the critics. Combining digital access with preservation belongs to the author. Search tools can speed retrieval is accepted by both in this invented passage. Now ask whether the author endorses eliminating originals: no. Cite the preservation rationale. This reading method fits LSAC passage analysis while leaving external archive-policy questions outside the exercise.

Source alignment: [L2/L3]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

Critics call archives obsolete; author supports search plus preservation. Identify author conclusion.

1. Mark the critic position.
2. Mark the author concession about speed.
3. Locate the preservation reason.
4. Author supports combining the two functions.

Mental model: Separate author judgments from positions the author reports.

Common trap: Assign every passage voice to the author.

16. Mixed arguments need a stem-specific review

Learning goal: Use a compact practice set to separate necessity, explanation and valid inference.

An argument can invite different tasks: identify an assumption, explain a discrepancy, or infer what follows. A statement that would strengthen the argument is not necessarily a required assumption. An explanation of how two facts coexist is not necessarily a conclusion entailed by them. Read the task word before reviewing options.

Original set: first, every approved application has a reference and Jo has no reference. Infer approval status. Second, prices fell while revenue rose; offer a possible explanation. Third, a survey of current visitors supports a claim about all prospective visitors; identify the needed bridge. Work each task in its own terms rather than use a single strategy for all three.

The first yields not approved by contraposition. The second may be explained by enough growth in quantity sold, but that explanation is not proven by the observations alone. The third requires some support for treating the sampled group as informative about the target group; assuming every potential visitor was surveyed is stronger than needed. The method changes with the stem.

For follow-up, alter the first case to Jo has a reference. Approval is now not established. Alter the second to quantity unchanged: lower prices alone cannot explain higher revenue in the simple price-times-quantity model. Alter the third to a properly limited claim about the actual respondents: the broad population bridge is no longer needed for that limited conclusion. Record the inferential step, not just the answer. These original questions are skill practice, not a complete LSAT section, a timing simulation or a score prediction.

Source alignment: [L2/L3]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

Approved → reference. Jo has no reference. Infer status.

1. Use the contrapositive.
2. No reference rules out approval under the premise.
3. Jo is not approved.
4. This does not require knowing other application details.

Mental model: Use a compact practice set to separate necessity, explanation and valid inference.

Common trap: Use a strengthen answer for an inference task.