The LSAT produces a single scaled score between 120 and 180, yet that four-digit number obscures a richer dataset that successful candidates learn to read with surgical precision. Your section-by-section performance — the gap between your Logical Reasoning and your Reading Comprehension, for instance — reveals which question families offer the largest score-improvement returns, how your study hours should be redistributed across preparation stages, and whether your current trajectory will close the gap to your target law school threshold. This article analyses LSAT section scoring mechanics, explains how section performance maps onto scaled-score growth, and outlines a diagnostic framework for converting raw section data into a targeted, high-efficiency study plan.
The architecture of the modern LSAT: how sections contribute to your score
The LSAT consists of four scored sections — Logical Reasoning (LR), Analytical Reasoning (AR), Reading Comprehension (RC), and a variable section that may be an experimental LR, AR, or RC — plus an unscored Writing Sample. Each scored section contains approximately 23 to 27 questions with a 35-minute time allocation, giving candidates roughly 85 seconds per question across the full test. The variable section does not affect your scaled score, but its presence is unknown during the exam, which means every section must be treated as scored.
The raw-score-to-scaled-score conversion applies uniformly across all sections during the equating process. In practical terms, this means that missing four questions in LR carries the same scaled-score penalty as missing four questions in RC, but the cognitive skill-sets required to close those gaps differ substantially between sections. Understanding this distinction is the first step toward converting your diagnostic data into a strategic study plan rather than a generic review schedule.
Reading Comprehension: diagnosing the passage-understanding versus question-handling divide
Reading Comprehension presents four passages of approximately 450 to 550 words each, with five to eight questions per passage. Questions fall into three broad families: main-point and purpose questions that probe global understanding; specific-reference and inference questions that require within-passage reasoning; and comparative passages that test the ability to synthesise and distinguish across two related texts. Performance in RC is frequently misunderstood as purely a reading-speed issue, but diagnostic data from full-length practice tests reveals that most candidates underperform due to a structural misinterpretation rather than a reading-rate constraint.
A candidate who consistently answers main-point questions incorrectly while performing adequately on specific-reference items is exhibiting a passage-level comprehension gap rather than a question-type weakness. Conversely, a candidate who grasps the passage structure accurately but struggles with inference questions is demonstrating a reasoning-gap pattern that requires targeted question-handling drills rather than additional passage practice. These two failure modes demand fundamentally different study interventions, and section-level score analysis is the only reliable method for distinguishing between them without conducting a full per-question audit of every practice test.
RC improvement also depends on passage-type recognition. Humanities passages (law, philosophy, cultural criticism) tend to employ abstract argumentation structures that reward the ability to identify the author's central claim before answering any question. Science passages, by contrast, often present structured experiments or empirical findings where the conclusion is embedded within a specific paragraph, making targeted reference questions more tractable once the candidate learns to locate those paragraphs with consistency. Science passages reward different reading strategies than humanities passages, and section-level analysis helps identify whether your RC weakness is concentrated in one passage type or distributed across all four.
Logical Reasoning: the largest score-improvement opportunity for most candidates
Logical Reasoning comprises two sections, each containing approximately 23 to 26 questions. The question families include assumption, flaw-in-the-reasoning, strengthener, weakener, inference, and principle-match items. LR is where the majority of LSAT score gains are realised during a well-structured preparation programme, and this is not an accident: LR questions test a skill-set that is both learnable and measurable, meaning that deliberate practice with quality explanations produces reliable score improvements within a predictable timeframe.
The critical distinction in LR preparation is between question-level mastery and section-level stamina. Most candidates who plateau in the 160-165 range have achieved question-level competence — they understand the logical structure of each question family — but struggle with the transition between questions within a section. This transition failure manifests as decreasing accuracy in the final five to seven questions of each LR section, which is a stamina and focus management issue rather than a logical-reasoning deficiency. Recognising this pattern changes the nature of the study intervention: the candidate does not need additional logical drills but rather section-length simulation practice combined with pacing-strategy refinement.
Within LR, assumption questions and flaw questions are the two families where candidates most frequently misidentify the logical force of the stimulus. An assumption question requires the candidate to identify the unstated premise that the argument's conclusion depends upon; a flaw question requires the identification of the specific reasoning error committed by the argument. Candidates who conflate these two question types — common among those who have not studied formal logic explicitly — tend to select answer choices that describe the argument's conclusion rather than the gap in its reasoning chain. Section-level data that shows a concentration of errors in assumption and flaw families points to a need for targeted formal-logic study, specifically the rules governing conditional reasoning and sufficient/necessary conditions, which are the conceptual foundation of both question families.
Analytical Reasoning: translating game-type familiarity into score gains
Analytical Reasoning, commonly known as logic games, presents four games per section with five to seven questions each. The games test the ability to derive implications from a set of conditional rules and to apply those implications across a variety of question scenarios. AR has historically been the section most amenable to rapid score improvement, because game-type recognition and diagramming fluency both improve substantially with systematic practice. However, the section also punishes candidates who attempt to rely on intuitive reasoning rather than systematic deduction, which makes structured game-type classification a high-priority study task for any candidate whose AR score lags behind LR or RC.
The primary diagnostic question for AR candidates is whether their errors cluster in setup questions (which test whether the candidate correctly derived the full set of allowable arrangements) or in hypothetical scenario questions (which test whether the candidate can correctly evaluate a specific conditional outcome against the game's rule set). Setup errors indicate an inadequate understanding of the game's rule structure, typically because the candidate drew an incomplete or incorrectly structured diagram. Scenario errors, by contrast, indicate either a failure to correctly apply a conditional rule to the hypothetical or an inability to hold all game constraints simultaneously when evaluating the scenario. Each failure mode demands a different corrective strategy: for setup errors, re-drawing diagrams from scratch without reference until the diagram becomes second-nature; for scenario errors, systematic substitution practice where each hypothetical scenario is tested against every rule before selecting an answer choice.
Game types follow predictable structural patterns — sequencing games, grouping games, hybrid games, and numerical distribution games — and each type has an optimal diagramming approach. Candidates who attempt to use a single diagramming method across all game types tend to underperform, because a diagram that efficiently captures a linear ordering may be entirely inadequate for a multi-group distribution problem. Section-level analysis that reveals consistently lower accuracy on specific game types directly informs which game-family drills should receive priority in the study schedule.
From section diagnostics to study-hour allocation: a decision framework
Converting section scores into an effective study plan requires ranking sections by two independent variables: current score deficit relative to your target, and improvement velocity — the rate at which focused practice produces measurable score gains in that section. Logical Reasoning typically exhibits the highest improvement velocity because the underlying logical skills are highly trainable and the question families are relatively consistent across test administrations. Reading Comprehension improvement velocity is more variable, depending heavily on how rapidly the candidate builds passage-type recognition and inference-habit modification. Analytical Reasoning improvement velocity is high for candidates in the early preparation stages but tends to plateau once game-type familiarity is established, because the marginal return on additional AR drills diminishes after the candidate has internalised the core diagramming strategies.
The practical implication is that study-hours allocation should not simply follow current score deficits. A candidate whose AR score is ten points below target but whose LR score is fifteen points below target might still achieve greater total-score improvement by prioritising LR study, because the LR improvement potential is higher even though the absolute gap is smaller. This counterintuitive logic is why section-level analysis is indispensable: it enables evidence-based study prioritisation rather than intuition-driven selection.
| Section | Primary failure mode | Recommended intervention | Typical improvement velocity |
|---|---|---|---|
| Logical Reasoning | LR structure confusion; stamina decline in final questions | Family-specific drill + section simulation pacing | High (8-12 weeks of focused study) |
| Analytical Reasoning | Incomplete game setup; conditional rule misapplication | Game-type classification + diagram fluency drills | High initially, plateau at ~165-168 |
| Reading Comprehension | Global comprehension gaps; inference habit errors | Passage-type strategy + inference identification drills | Moderate (12-20 weeks) |
Common pitfalls and how to avoid them
The most consequential error candidates make when using section-level data is aggregating LR-1 and LR-2 performance without distinguishing between them. These sections are structurally identical, but the adaptive difficulty mechanism in recent LSAT administrations means that the second LR section may contain more difficult stimuli if the candidate performed strongly in the first. Treating LR as a single undifferentiated category masks the possibility that your LR-1 errors are concentrated in one question family while LR-2 errors are concentrated in another, which would indicate a preparation strategy that needs to address both question families across both sections rather than treating one section as the problem.
A second common pitfall is interpreting section scores as fixed rather than relative. Candidates who score 167 in RC and 160 in LR frequently focus their remaining study time on AR because their RC score is already stronger, but this reasoning misunderstands how the scaled-score distribution works. A three-point gain in LR from 160 to 163 contributes more to your overall scaled score than a three-point gain in AR from 163 to 166, because score gains at the lower end of the distribution are worth more in raw-scaled terms than equivalent gains at the upper end. Studying the section where you are currently weakest may not be the most efficient path to total-score improvement if that section has a lower improvement ceiling than a section where you are moderately strong.
A third pitfall is conflating question familiarity with question competence. A candidate who has seen 200 assumption questions and answered them at 75% accuracy has developed familiarity with assumption questions, not mastery. Genuine mastery is demonstrated by the ability to correctly identify the argument's unstated premise under time pressure across a full section of 23 to 26 LR questions. Section-level analysis that reveals consistent accuracy in individual questions but declining accuracy in the final ten questions of the section is diagnostic of a stamina deficit, not a competence deficit, and stamina deficits cannot be resolved by additional question-type drilling. Only full-length section simulations with enforced pacing can remediate stamina-based performance degradation.
Setting section-level targets from your law school goal
Your target total LSAT score is derived from the median and 75th-percentile scores of your target law schools, but translating that total into section-level targets requires accounting for the distribution of scaled-score increments across the range. A target of 170 — sufficient for competitive admission at many top-tier programmes — translates to an average of approximately 42.5 raw correct answers per section across the four sections, accounting for the non-linear relationship between raw scores and scaled scores at different points in the range. This is an approximation because equating varies across test administrations, but it provides a useful benchmark for section-level target-setting.
Rather than targeting equal scores across all sections, candidates should aim for a distribution that reflects their natural cognitive strengths. A candidate with a humanities undergraduate background and strong reading comprehension will typically find RC the most tractable section to push toward the upper end of the range, while a candidate with a quantitative background may find LR's explicit logical structure more naturally aligned with their analytical training. Using your section-level diagnostic data to identify which sections are closest to your ceiling and which have the most remaining improvement potential enables you to construct a realistic section-score target that, when combined, aligns with your overall score goal.
Monitoring section-level progress across your preparation timeline
Effective preparation tracks section scores across multiple full-length practice tests, not just the aggregate total. A candidate whose total score has remained at 163 across three practice tests may nonetheless be making significant progress that is invisible when looking only at the aggregate figure. If their LR has improved from 160 to 165 while their AR has declined from 167 to 162, the total remains unchanged but the underlying diagnostic picture has changed substantially, and the study plan should reflect that change rather than treating the candidate as static.
The recommended cadence for full-length section-level analysis is one complete practice test every seven to ten days during active preparation, with section subscores tracked in a spreadsheet that includes both the scaled-score estimate for each section and the raw correct-answer count. Over a sixteen-week preparation cycle, this produces approximately eight to ten data points per section, enabling the identification of trend lines rather than single-test outliers. A section that shows consistent improvement across six consecutive tests is likely to continue improving with maintained study; a section that fluctuates without trend suggests an unstable skill foundation that may require diagnostic re-evaluation.
Scores that plateau for three or more consecutive tests indicate that the current study method has reached its effectiveness ceiling for that section. When a plateau occurs, the appropriate response is not to increase the volume of the same type of study but to change the study modality. For LR plateaus, this might mean shifting from question drills to formal-logic study; for AR plateaus, it might mean working through the most difficult game types in the section's upper difficulty range rather than continuing to drill familiar game structures; for RC plateaus, it might mean reducing passage practice in favour of targeted inference-habit training using individual stimulus-response pairs isolated from full passages.
Conclusion and next steps
Your LSAT section scores are not merely a breakdown of your total — they are a strategic diagnostic tool that, when interpreted correctly, transforms the allocation of your preparation hours from guesswork into evidence-based decision making. The candidate who understands that their LR performance deficit is concentrated in assumption questions and caused by conditional-reasoning confusion, rather than distributed across all question families and caused by generic logical weakness, can target the precise conceptual gap that is costing them points. This level of diagnostic precision is available from every full-length practice test, but it requires systematic section-level tracking rather than total-score monitoring alone.
TestPrep's complimentary diagnostic assessment offers a structured starting point for candidates seeking to convert their section-level performance data into a prioritised study plan aligned with their law school admission goals.