CAT Raw Score vs Scaled Score vs Percentile Explained

Raw score comes from your responses under the marking scheme. CAT then normalises across test forms/slots to produce scaled section and total scores. Percentile is your rank position relative to test takers based on scaled scores.

Start with raw score

Raw score is the closest thing to exam arithmetic. Under the usual CAT scheme, a correct response adds marks, an incorrect MCQ can reduce the score, and incorrect TITA/non-MCQ responses do not attract the MCQ penalty. For example, If a candidate has 20 correct MCQs, 5 wrong MCQs and 2 correct TITA questions, illustrative raw score = (22×3) − 5 = 61. The eventual scaled score can differ after normalisation.

Why CAT then uses a scaled score

CAT is conducted in multiple sessions. The official scoring/equating process exists so that candidates are compared more fairly when different forms or slots are not identical in difficulty. The process adjusts score distributions and reports scaled scores for VARC, DILR, QA and the total. It is the scaled score, not a simple self-calculated raw total, that underpins the reported percentile.

Percentile is a rank position, not a percentage of marks

A 99 percentile does not mean you scored 99% of the total marks. It means your scaled performance placed you ahead of roughly 99% of the relevant candidate pool under CAT's percentile calculation. That is why two adjacent scaled scores at the top can sometimes correspond to meaningful percentile differences.

What should a student do with this?

Before the exam, use raw-score targets for mock analysis because you can control correct answers, negative marks and selection. After the official scorecard, use the scaled score and reported percentile for admission decisions. Do not try to reverse-engineer a future exact percentile from a raw score weeks before the exam.

Raw score and scaled score are not interchangeable, and a raw score cannot be converted to an exact future percentile before the cohort/slot distributions are known.

The practical rule is: mock with raw-score discipline, plan with score bands, and make admission decisions from the official scaled score/percentile.