SNAP 2026 Exam Pattern and Marking: 60 Questions, With a Calculator

The total did not change: SNAP 2025 was also 60 questions, 60 marks and 60 minutes. What changed is where the marks sit. General English lost five questions and Reasoning lost five, and those ten became the Ethics, Morality and Values section. Quantitative stayed at 20. So a candidate whose plan was built on the 2025 paper now has one-sixth of the paper in a section that did not exist last year, and a smaller English section than before.

The paper is objective and in English. Each question offers four responses and you choose one. The bulletin says timings for each test will be printed on the admit card, so there is no published start time yet. Candidates registered as Persons with Disabilities get 20 minutes of compensatory time for every 60 minutes; SIU does not provide a scribe, the candidate arranges one, and the scribe must not have studied beyond Class XII. Calculators, smart watches and other electronic devices are not allowed. If an error is found in a question paper, a Scrutiny Committee decides after the tests, and the bulletin says scores published by SIU are final and not open to appeal or revaluation.

The same arithmetic, worked for four candidates. These illustrate the marking only; they are not predictions of any percentile.

Two things stand out. Four wrong answers cancel one right answer, so a candidate who answers every question and gets a quarter of the attempted ones wrong loses about six marks to the penalty. And the scaled figure matters because the merit list uses it: SIU takes "SNAP Score (out of 60, scaled down to 50 marks)" as half of the 100-mark merit score. The 5/6 factor is our arithmetic; the bulletin does not print the formula.

With four options, a blind guess is right one time in four. One right answer (+1) against three wrong ones (3 × -0.25 = -0.75) leaves +0.25 over four guesses, or +0.0625 per guess. Every option you can rule out raises that value. The table below is our arithmetic on the official marking, not SIU advice.

On average, then, answering beats leaving a question blank whenever your chance of being right is above one in five, which a genuine four-option guess always is. But the official SNAP blog on the new section gives the opposite advice for uncertain questions: "Candidates should avoid uncertain guesses because every incorrect answer attracts 25% negative marks. Eliminate clearly unsuitable choices and answer only when the reasoning is strong enough." Both can be true at once. Expected value is an average over many questions; on any one question a guess can cost you a quarter mark, and in a 60-minute paper the time spent on a question you cannot solve is itself a cost. Our reading: eliminate first, answer when you have ruled out at least one option, and use your mocks to see whether your guesses actually pay.

You may book up to three SNAP tests. The bulletin says: "If a candidate appears for more than one test, the higher score will be considered for the final percentile calculation and there will be no normalization." The home page puts it as "Best of 3 Attempts. Your best score counts. No normalisation".

The same bulletin paragraph goes on to say that a statistically tested process frames unique questions under each topic, that this "will ensure normalisation of the spread of questions across sections", and that "the question papers will be normalised for the level of difficulty across all three tests". So one official paragraph says both "no normalization" and "normalised for the level of difficulty". Our reading is that the first sentence is about scores (your higher raw score is used as it is, without statistical scaling between tests) and the last sentences are about how papers are set (made comparable in difficulty in advance). The bulletin does not say so in terms, and it adds that no correspondence on this will be entertained. SIU publishes no percentile formula, no score-to-percentile table and no count of candidates.

The general-awareness section went after 2016, the paper was halved in length and time after 2019, and in 2026 an ethics section arrived. One thing has not moved in any document we have: a wrong answer costs a quarter of the marks the question carries. That is why the 2019 figure of -0.5 existed at all. It was a quarter of a 1.5-mark question, not a harsher rule.

Sixty questions in sixty minutes is one minute each on average, but the sections do not cost the same. Reading passages and data sets take longer per question than a short vocabulary item or an ethics scenario. The official preparation blog suggests practising different section orders in mocks before fixing an order for the test, answering clear questions first, marking harder ones and returning when time permits. Because there is no sectional timer, a section you leave for the end can be squeezed; a plan that fixes a rough minute budget per section, tested in mocks, protects all four.

A note on our own content: the SNAP exam record on ProSyllabus still shows the older +1.5 / -0.5 marking and a three-section syllabus without the ethics section. This guide follows the SNAP 2026 bulletin; the record is being corrected separately.