Normalization is a statistical adjustment used when an exam is held across multiple sessions with different difficulty levels, so no candidate is unfairly advantaged or disadvantaged by which session they sat.
Large-scale exams sometimes run across multiple days or shifts to accommodate all candidates. If one shift's question paper happens to be harder than another's, raw scores wouldn't be directly comparable — a normalization formula adjusts scores to account for that difference in difficulty.
Normalization typically compares each candidate's performance to the overall performance of everyone in their specific session, then converts that into an equivalent score on a common scale — so a raw score of 80 in an easier shift might normalize to a lower equivalent score than an 80 in a harder shift.
The exact formula varies by exam and examining body, and is usually published in the exam's official information bulletin — candidates preparing for a specific exam should check that source for the precise method used.
Last reviewed: September 2026