CEETA PG 2026 Exam Pattern

The revised 100-question structure

The reviewed 2026 pattern information gives 100 objective questions in a two-hour offline examination. The distribution is 10 questions in Engineering Mathematics, 15 in Basic Engineering and Sciences, and 75 in the selected discipline. Some secondary tables still show older or internally inconsistent distributions. Preparation should therefore use the revised 10–15–75 structure, while candidates must always follow the actual booklet instructions.

The reported marking scheme awards one mark for a correct answer and deducts one quarter of a mark for an incorrect answer. A blank answer does not earn a mark. This structure makes accuracy important, but it does not imply that a candidate should spend equal time on every question. Some conceptual items may take seconds, while a numerical item may require several linked steps.

Seventy-five of the hundred questions belong to the selected discipline. Common Mathematics and Basic Engineering remain worth preparing, but they should not consume most of the revision time simply because a generic study guide discusses them first. The candidate's undergraduate technical foundation is central to this examination's preparation.

Raw marks and an attempt example

Consider an original practice result with 68 correct answers, 20 incorrect answers and 12 blanks. Under a plus-one and minus-one-quarter scheme, the raw total is 68 minus 5, which equals 63. This calculation is a practice illustration. It is not a prediction of a percentile, a rank or an admission cutoff.

Now consider another practice attempt with 63 correct answers, 4 wrong answers and 33 blanks. Its raw total is 62. The first candidate attempted many more questions but finished only one mark ahead. The comparison shows why attempted-question count alone is a poor progress measure. Record correct, incorrect and unanswered items separately when reviewing a mock.

There is also a time cost to uncertainty. Spending six minutes on a question with no clear method can prevent three solvable questions from being reached. A useful first pass secures direct answers and identifies questions that need longer working. The second pass should prioritise those with an identifiable method, rather than returning automatically to the earliest difficult item.

For four equally plausible options, a random selection has a theoretical expected score of one sixteenth of a mark under this scheme. That mathematical average is not a promise for a short examination and says nothing about the time spent. Decisions should consider whether options can be eliminated, whether the premise has been understood and whether other questions offer a better use of the remaining time.