GMAT – Foundation & Exam Strategy
Topic: Marking Scheme
Exam: Graduate Management Admission Test (GMAT)
Section: Quantitative, Verbal, Integrated Reasoning, Analytical Writing Assessment (AWA)
1. Overview of Points Distribution
| Section | Total Points | Sub‑sections | Scoring Formula |
|---|---|---|---|
| Quantitative | 60 | Problem Solving (PS) – 31 questions Data Sufficiency (DS) – 29 questions |
Raw score → scaled score 200‑800 (linear interpolation) |
| Verbal | 60 | Reading Comprehension (RC) – 8 questions Critical Reasoning (CR) – 13 questions Sentence Correction (SC) – 23 questions |
Raw score → scaled score 200‑800 |
| Integrated Reasoning | 0‑8 points | 8 questions (multiple‑choice, graphics, multi‑source) | Each correct answer → 1 point (no penalty) |
| Analytical Writing Assessment | 0‑6 points | 1 essay | Two trained raters score 0‑6 in half‑point increments; scores averaged |
Total possible score: 800 (Quant + Verbal) + 8 (IR) + 6 (AWA) = 814
Note: Only Quant and Verbal are combined for the overall GMAT score (800). IR and AWA are reported separately.
2. Detailed Scoring Mechanics
2.1 Quantitative & Verbal
| Step | Description | Example |
|---|---|---|
| Raw Score | Count of correct answers (no negative marking). | 27 correct out of 31 PS questions → Raw = 27 |
| Scaled Score | Non‑linear mapping using item‑response theory (IRT). The GMAT committee calibrates test‑equating across administrations. | Raw 27 → Scaled 600 (approx) |
| Equating | Ensures comparability across test dates. A raw score of 27 on a “hard” test equates to the same scaled score as a raw 29 on an “easy” test. |
Key Points
- No penalty for incorrect answers.
- Answering all questions is mandatory; leaving blanks gives no advantage.
- Item difficulty varies; a correct answer on a harder question contributes more to the overall ability estimate than an easy one, but this is handled internally by the equating process.
2.2 Integrated Reasoning
- Scoring: 1 point per correct answer.
- No penalty for wrong answers.
- Scoring vs. Percentile: For each test date, scores are converted to percentiles (e.g., 5th, 50th, 95th).
- Use in admissions: Some schools use the IR score as a separate indicator of analytical ability.
2.3 Analytical Writing Assessment
- Scoring: Each essay is scored 0–6 by two raters; final score is the average (rounded to nearest half).
- Calibration: Raters are trained to maintain inter‑rater reliability (≈0.88).
- Skills Assessed: Ability to construct a coherent argument, use evidence, and write clearly.
3. Tricky Points & Common Misconceptions
| Misconception | Reality | Why It Matters |
|---|---|---|
| Answering fewer questions can improve score. | No – All questions count equally; leaving blanks never boosts your scaled score. | Wrong strategy causes wasted time. |
| Difficulty of a question changes the point value. | No – Each correct answer contributes the same raw point; difficulty only affects the probability of answering correctly. | Misunderstanding can alter pacing strategy. |
| Wrong answers lower the score. | No – There is no penalty; only correct answers increase raw score. | May lead to over‑cautious behavior. |
| IR score can be ignored. | Not always – Some business schools weigh IR heavily, especially for analytical reasoning. | Ignoring IR may hurt admission chances. |
| AWA score is irrelevant. | Depends on school – Many top B-schools consider AWA, especially for leadership potential. | A weak AWA can offset a high Quant/Verbal score. |
4. Derivation of Scaled Score (Illustrative)
The GMAT uses Item Response Theory (IRT) to map raw scores to scaled scores. While the exact formulas are proprietary, the general idea is:
- Estimate ability (θ) of test taker from pattern of correct/incorrect responses.
- Equate across test versions: For each raw score R, determine a scaled score S such that the percentile rank is consistent across administrations.
Mathematically: [ S = a \cdot \Phi^{-1}\left(\frac{R + 0.5}{N}\right) + b ]
- ( \Phi^{-1} ) – inverse cumulative normal distribution.
- ( N ) – total questions.
- ( a, b ) – calibration constants (derived from field data).
Important: The mapping is non‑linear; small differences in raw score at the high end can produce larger changes in scaled score.
5. Study Tips Focused on Marking Scheme
| Tip | Rationale |
|---|---|
| Answer all questions | No penalty for incorrect answers. |
| Time‑boxing | Allocate ~1.2 min per Quant, ~1.5 min per Verbal question. |
| Track “known” vs. “unknown” | If you’re confident in a question, answer it immediately; otherwise mark and return. |
| Use process of elimination | For Verbal, eliminate options to improve odds; for Quant, skip if stuck and come back. |
| Practice with full‑length tests | Builds stamina and familiarity with marking nuances. |
| Review IR practice | Even though scoring is linear, understanding the format boosts confidence. |
| Draft AWA | Follow the 10‑minute structure: 2‑minute intro, 5‑minute body, 3‑minute conclusion. |
6. Hard Practice Question (Quantitative)
Problem
A company sells two types of widgets: A and B.
- Widget A costs $3 to produce and sells for $5.
- Widget B costs $4 to produce and sells for $7.
The company’s total production cost for a day is $1,200, and its total revenue for that day is $1,800.
Let (