NEET Leaks and How to Prevent it. - By Group Captain Johnson Chacko

Date:- 30th July 2026

Introduction

My maid sold her land to pay 3 Lakhs for NEET coaching for her daughter. When the paper leaked, she almost committed suicide. About 20 Lakh students appeared. The beneficiaries get 60,000 Crores. Who are the beneficiaries in the food chain, and will all of them be held to account?

Question paper leaks need to be prevented at all costs. What encourages paper leaks is the ecosystem that is built around the examinations.

● The causes for paper leaks are:

● The need for candidates to qualify for professional courses.

● The limited seats that are available.

● Many people think a common paper is inherently fairer. But from a measurement perspective, that is not always true.

● The main advantage of a single common paper is simplicity and transparency:

-Everyone faces the same questions.

-No need to estimate paper difficulty.

-No controversies about the normalization formula.

-Ranks follow directly from marks.

This is exploited by interested parties and coaching centres that mushroom.  They need to show that candidates trained by them are better. Getting the paper in advance and teaching the candidates answers to those questions is a sure shot way of enhancing their demonstrated performance leading to their popularity and their income in future. When this question paper is leaked it compromises the entire process.

Aim of NEET:

NEET is the entrance test for medical and related courses. Candidates receive marks, but admissions are made based on the rank (All India Rank /State Rank) derived from those marks.

Suggested Solution:

If different candidates receive different question papers, exam authorities can use normalization or equating methods to adjust scores so that candidates are not disadvantaged by getting a harder paper. The idea is to estimate what a candidate would have scored had everyone taken papers of equal difficulty. Capability of the candidate and degree of difficulty are the variables.

With NEET's very large candidate population, NTA could administer multiple equivalent question papers and use modern psychometric methods (Standard Deviation and Mean) to normalize scores.

However, there are challenges

● The normalization model must be statistically robust.

● Different papers must be carefully calibrated beforehand.

● Public trust in the process is essential.

● Litigation and disputes may arise if students feel the adjustment was unfair.

Marks from different papers can be normalized and ranked statistically, and this is a well-established practice. NEET currently relies on a common paper largely to avoid the complexity and perceived unfairness that can accompany normalization.

With a very large candidate population and truly random assignment of candidates to test papers, statistical normalization can produce rankings that are practically very fair. If each paper is attempted by a sufficiently large and randomly selected sample of candidates, then the ability distribution of candidates in each paper should be approximately the same.

Results:

● Differences in the mean score across papers are mainly due to paper difficulty.

● Differences in the spread (SD) are mainly due to paper characteristics.

● Therefore, normalization can adjust scores onto a common scale.

Suppose NEET has 2 million candidates.

Assume:

● 10 different papers.

● 200,000 candidates per paper.

● Random Allocation

By the Central Limit Theorem and the Law of Large Numbers, each group's underlying ability distribution should be nearly identical.

The challenge is not the mathematics itself. The challenge is the assumptions.

The conclusion holds if:

● Candidates are randomly assigned.

● No paper leaks occur.

● All papers test the same syllabus and cognitive level.

● Candidate populations are genuinely comparable.

If these assumptions hold, then normalization is difficult to criticize statistically.

The Process:

● The candidate should be allocated the paper randomly from the server, once he checks into the examination hall. The admit card should have his photo.

● The seat allocation should be allocated in a manner that no adjacent seats are allocated for NEET to prevent copying.

● The exam centres should have CCTV coverage of all desks.

● Time is controlled by the system and it logs off when the time is over.

● Marks may be displayed when the candidate logs off and may be sent by email.

● Candidate should not be allowed anything other than a pen/pencil. Rough paper should be provided by the exam centre and taken back after the exam.

The advantages of this process:

● The exam s not dependent on a single date and time. It can be spread over a longer period. Candidates can choose their date and time at the time of registration.

● Coaching centres must really coach the entire syllabus.

● A leaked paper will be of no value as the candidate may not get that paper for the test. If all 100 sets of papers are leaked and a candidate studies them all, then he deserves to pass as that would cover the entire syllabus.

● Use of Air Force helicopters won’t be needed.

● Normalisation of answers takes place centrally and cannot be manipulated as access to the algorithm can be restricted.

Conclusion:

The issue is often public acceptance, not statistical validity. For a massive examination like NEET, if candidates are randomly distributed across multiple equivalent question papers, then normalization based on score distributions can produce a rank order that is statistically justifiable and likely very close to the candidates' true relative abilities.

The real debate is not whether such normalization can work. It certainly can. The debate is whether examination authorities and candidates trust the assumptions and methodology enough to accept the resulting ranks.

In fact, many psychometricians would argue that with millions of candidates and proper test design, a normalized multi-paper system can be as fair as, and sometimes fairer than, a single-paper system. The resistance is often administrative, legal, and perceptual rather than mathematical.

All the stake holders including the candidates and their parents need to be adequately sensitised before these changes are made. JEE Mains does not seem to have this problem.

I have normalised the marks obtained by Flight Cadets of flying branch, of different streams (fighters, transports, helicopters and navigators) from different academies where the syllabus, aircraft flown etc. are unrelated to each other, to achieve a common merit for allocation of service numbers in the IAF, statistically. The same was done for Aeronautical Engineers of Mechanical and Electronics sub-branches. It continues even after 25 years.