From Moaz Nair
When hundreds of thousands of Malaysian students wait anxiously for their SPM results, or their peers across the causeway scan the O-Level boards, they are told that they are being evaluated purely on merit.
Educational syndicates proudly market these high-stakes gauntlets as “criterion-based” — an absolute measure of what a student knows, independent of anyone else.
But behind closed doors, a very different statistical reality plays out.
Contrary to what has been assumed to be criterion-based, Cambridge IGCSE and O-Level systems do not use fixed score thresholds. Instead of a rigid “90% equals an A*” rule, Cambridge adjusts grade boundaries after every exam series based on the difficulty of the specific test papers.
While the system evaluates students against fixed learning criteria, it incorporates statistical screening to ensure fairness; if an exam paper is exceptionally difficult, the mark required for a top grade drops.
This flexible approach ensures that a grade reflects the same level of achievement from year to year, without placing strict quotas or caps on how many students can earn an A* (A-star).
Almost all major examination boards — including our own Examination Syndicate (Lembaga Peperiksaan) and Singapore Examinations and Assessment Board (SEAB) — quietly layer norm-referenced adjustments on top of raw marks.
Colloquially known as “grading on a curve” or “standard fixing”, this hidden mathematical machinery ensures that an “A” earned in an unexpectedly brutal exam year holds the exact same currency as an “A” from an easier year.
While it keeps national systems stable, this invisible balancing act is reaching a dangerous tipping point, triggering a global reckoning over student burnout and economic reality.
How the curve secretly rules the world
Different countries mask or embrace this statistical architecture in unique ways, shaping how students progress into universities and the workforce.
Malaysia’s SPM operates on a “standard fixing” model — a criterion-related, norm-referenced hybrid.
While strict subject rubrics exist, the exact raw score thresholds required to secure an A+, A, or A- shift dynamically year-by-year based on the national cohort’s overall performance curve.
In the UK, the Office of Qualifications and Examinations Regulations (Ofqual) uses a policy called “Comparable Outcomes”.
By mapping the current batch of A-Level students against how they performed years ago on their childhood exams, they mathematically predict what percentage should get top grades today, adjusting the grade boundaries to match.
Singapore, in collaboration with Cambridge Assessment, maintains a fiercely competitive system where cohort distributions fine-tune final grades.
This ensures consistent standards across different generations of test-takers competing for fixed tertiary slots.
South Korea’s Suneung, a high-stakes eight-hour college scholastic ability test, completely discards raw scores for standard percentiles. Under Suneung only the top 4% of the entire country can ever secure a Tier 1 grade.
Similarly, Australia scales different subjects against one another to output the Australian Tertiary Admission Rank — a pure percentile rank from 0.00 to 99.95.
In the US, the College Board uses a process called “equating” for the college admission test, known as “SAT”. Instead of scaling you against the people in the room on test day, they map your performance against a universal “norm group” of historical test-takers to ensure a score of 700 means the same thing today as it did a decade ago.
A fundamental trade-off
The tension between norm-referenced and criterion-referenced systems comes down to a fundamental trade-off between institutional fairness and individual well-being.
On one hand, norm-referenced grading acts as a safety net. If an exam board accidentally sets an impossibly difficult Mathematics paper, the curve automatically drops the required mark for an “A”, protecting students from being penalised for the year they happened to be born.
It also provides universities with an incredibly clean, efficient filter to select top-tier applicants.
However, the dark side is a toxic, zero-sum environment. It breeds a culture where a peer’s failure directly benefits your grade.
Worse, it masks systemic educational failures. If an entire generation’s literacy or mathematical mastery drops due to prolonged disruptions, standard-fixing can hide the rot by keeping the national percentage of A’s completely stable.
Intentional or not, using standard-fixing as a political shock absorber risks severing the link between a student’s certificate and their actual, real-world proficiency.
When the curve is adjusted to ensure pass rates remain palatable, it acts less like an objective academic yardstick and more like cosmetic anaesthesia, hiding structural damage under a layer of predictable percentages.
Ultimately, centralised exam boards face an impossible mandate: they are asked to be absolute measuring tapes of human knowledge, yet they must function as traffic cops for a finite number of university seats and corporate slots.
By choosing the stability of the curve over the raw truth of student capability, our educational systems have sacrificed individual transparency for macro convenience.
Until we reconcile this contradiction, we will continue to subject hundreds of thousands of students to a game where the rules are hidden, the goalposts move, and a peer’s failure remains the secret ingredient to one’s own success.
The bell curve may keep the system level, but it is flattening the very minds it claims to measure.
Moaz Nair is an FMT reader.
The views expressed are those of the writer and do not necessarily reflect those of FMT.