🧑🎓 NEET PG – Foundation & Exam Strategy
📌 Topic: Marking Scheme
This note is written for students preparing for NEET PG who want an MIT‑level understanding of the marking system, why it matters, and how to exploit it strategically.
1. The “First Principles” of the Marking Scheme
| # | Feature | Rationale (from exam‑design perspective) |
|---|---|---|
| 1 | Answer Types | Single‑best answer (SBA) – 1 point per correct answer, 0 for wrong. Multiple‑choice (MC) – 1 point per correct, 0 for wrong. True/False (TF) – 1 point per correct. |
| 2 | Negative Marking | -0.25 for each wrong answer (SBA, MC, TF). No penalty for omitted questions (blank). |
| 3 | Question Weight | Each question carries equal weight (1 point). No “bonus points.” |
| 4 | Answer Entry | Paper‑based, with a fixed number of options (typically 4). |
| 5 | Scoring Formula | [ |
| \text{Score} = \sum_{i=1}^{N}\bigl( \mathbf{1}_{\text{correct}i} - 0.25,\mathbf{1}{\text{wrong}_i}\bigr) | ||
| ] where (N=180) (current NEET‑PG). |
Why this matters:
- Uniform weight → every question is equally valuable.
- Negative marking → forces risk‑calculation.
- No partial credit → maximises the need for certainty.
2. Probabilistic Analysis of the Marking Scheme
Let
- (p) = probability of answering a question correctly by guessing.
- For 4‐option MC, (p = 0.25).
- For 2‑option TF, (p = 0.5).
- (q = 1-p) = probability of an incorrect answer.
2.1 Expected Value of a Guess
[ E[\text{guess}] = 1\cdot p - 0.25\cdot q ]
Example (4‑option MC):
[
E = 1(0.25) - 0.25(0.75) = 0.25 - 0.1875 = 0.0625
]
So a blind guess still yields a positive expected value (≈ 0.06).
Edge‑case: If a question had more than 4 options, the expected value could become negative.
[
p = \frac{1}{k};;\Rightarrow;;E = \frac{1}{k} - 0.25\left(1-\frac{1}{k}\right)
]
Set (E=0) → (k = \frac{1}{0.25} = 4).
Thus any question with ≥ 5 options would discourage guessing.
2.2 Expected Score After (n) Guesses
[ E_{\text{total}} = n \times E[\text{guess}] ] Because each guess is independent, the linearity of expectation holds.
Implication: Even if you are completely clueless, you gain an average of 0.06 points per question, so for 30 guesses you expect +1.8 points.
3. Decision Theory: When to Guess, When to Skip
Define:
- (C) = confidence in a particular answer (0 ≤ C ≤ 1).
- (P_{\text{correct}} = C) (assuming only one answer is correct).
3.1 Expected Gain of Answering vs Skipping
- Answering:
[ E_{\text{ans}}(C) = C \cdot 1 - (1-C) \cdot 0.25 = 1.25C - 0.25 ] - Skipping:
[ E_{\text{skip}} = 0 ]
Answer if (E_{\text{ans}}(C) > 0) →
[
1.25C - 0.25 > 0 ;;\Rightarrow;; C > 0.2
]
Thus: If you have more than 20 % confidence that your answer is correct, you should answer.
Edge Case: If the question is worth 1 point but the cost of a wrong answer is higher (e.g., in some other exams), the threshold changes.
4. Advanced Strategies Derived from the Marking Scheme
| Strategy | Under What Conditions | Why It Works |
|---|---|---|
| A. “80 % Rule” | You’re ≤ 80 % through the paper and still have time to scan the remaining 20 % of questions once more. | Allows you to fill in high‑confidence guesses and regroup uncertain questions. |
| B. “Confidence Bands” | Rank all questions by your confidence (high → low). | (1) Answer all above‑20 % confidence. (2) Skipping those with < 20 % leaves you with fewer unanswered questions, preserving time for later. |
| C. “Negative‑Marking Counter” | Keep a mental counter of how many wrong answers you suspect you have. | If you exceed a safe threshold (e.g., > 40 wrong answers), you might switch to a “guess‑all” mode to recover value. |
| D. “Time‑Allocation” | 200 min ≈ 1.11 |