I was sitting at a high-school graduation when the honours began.
One student graduated with distinction.
Then another.
Then several more.
After a while, I started to wonder whether I had accidentally attended the graduation ceremony for a small nation’s entire gifted population.
My first thought was the obvious one:
They cannot all be above average.
Then I realized I was already making a mistake.
A school is not required to distribute achievement like height. If every student can demonstrate the required knowledge, every student can earn a high mark. A driving examiner does not have to fail thirty per cent of competent drivers to preserve the meaning of the licence.
Grades are supposed to be closer to a standard than a ranking.
So the interesting question is not:
Why were so many students honoured?
It is:
What did the honour tell me they were ready to do next?
An A can be common and still mean something
The bell curve has a strange hold on us.
If too many people receive high grades, we suspect somebody has been generous. If too many fail, we suspect the teaching. Somewhere between the two, we imagine, lies the honest distribution.
But there is no natural law requiring mastery to form a bell curve.
If a class begins at different levels, receives good teaching and is allowed to improve, its final results may bunch near the top. That could be a success.
The difficulty begins when the same symbol is asked to perform several jobs:
- tell a student what they understand;
- certify that they completed a course;
- rank them against other applicants;
- predict whether they can handle the next course;
- help a university choose among more qualified people than it has seats.
Those jobs are related.
They are not identical.
A mark can be perfectly fair inside one classroom and still be difficult to compare with a mark from another.
The number has lost its unit
Imagine two students applying to engineering.
Both earned 93 per cent in Grade 12 mathematics.
One teacher built assessments around unfamiliar problems, cumulative exams and limited retakes. Another used smaller units, generous corrections and assignments that closely resembled the practice material.
Both students may have worked hard.
Both teachers may have followed legitimate policies.
Both 93s may accurately summarize performance under those conditions.
But “93” is not a complete measurement unless we also know the instrument that produced it.
The missing unit
We would never compare 93 kilometres with 93 miles and complain that the numbers were misleading.
We would ask for the unit.
Grades arrive with a course name, but much of the unit—the difficulty, assessment design, support, revision policy and local standard—remains invisible.
Is this grade inflation?
Possibly.
But the phrase is often doing more work than the evidence.
In 2026, assessment researchers Christopher DeLuca and Michael Holden noted that their search found no published Canadian elementary or secondary data documenting post-2021 grade inflation. They also pointed out that rising admission averages can have other causes, including stronger applicants, improved teaching and competition for limited programs. (Queen’s Faculty of Education)
That does not prove inflation is absent.
It means I should not turn a surprising graduation ceremony into a national trend line.
There are at least four different possibilities:
- students are learning more;
- assessments or grading practices have become more generous;
- competitive programs are attracting stronger pools;
- the label “honours” has become more common even if the underlying grades have not.
These can all happen at once.
The graduation gave me an observation, not a diagnosis.
Canada mostly asks grades to travel alone
Canadian undergraduate admission is decentralized. Requirements vary by province, university and program, but secondary-school course grades carry much of the load.
Ontario applicants, for example, generally present six Grade 12 university- or mixed-level courses, and universities receive the reported grades through the Ontario Universities’ Application Centre. Selective programs may add essays, interviews or other supplemental material. (OUAC application guide)
There is no single national university entrance examination that recalibrates every applicant against the same tasks.
Alberta is a partial exception worth noticing. In diploma-exam subjects, a student’s final course mark combines the school-awarded mark and a common provincial exam, currently weighted 70 and 30 per cent. (Alberta diploma exams)
That does not make Alberta’s system automatically fair or complete. One exam cannot measure every worthwhile capability. High-stakes exams create their own distortions, and a common test may privilege students with better preparation for its format.
But it does something locally awarded grades cannot do alone.
It creates a shared reference point.
The admissions arms race is rational for almost everyone
Suppose seats in a program are scarce and grades are the main admission signal.
Students have a reason to protect every percentage point.
Parents have a reason to seek courses, schools and policies that produce competitive marks.
Teachers have a reason to avoid disadvantaging their students relative to other schools.
Schools have a reason to celebrate outcomes that make families want to attend.
Universities raise cut-offs because the applicant pool arrives with higher numbers.
Nobody has to conspire.
A signal under pressure
This resembles the problem in why every metric eventually learns to lie.
Once a measurement controls access to something valuable, people adapt to the measurement.
The adaptation may include harder work and better teaching.
It may also include easier assessment, strategic course selection and pressure against low marks.
The metric cannot tell us which kind of adaptation produced it.
Then the student meets calculus
Darja Barr’s 2019 University of Manitoba dissertation examined the relationship between Grade 12 pre-calculus and first-year calculus performance from 2001 to 2015.
The study found a growing disconnect over time, particularly among students who did not perform well in university calculus. Adding an ALEKS mathematics placement assessment improved prediction beyond high-school grades and demographic variables. (Barr, 2019, Theses Canada)
This is one dissertation at one university, not a verdict on Canadian schools.
Still, the pattern raises an awkward possibility.
Perhaps the high-school mark and the university course are measuring overlapping but increasingly different things.
The university discovers the difference when the student writes the midterm.
By then:
- the admission decision has been made;
- tuition has been paid;
- the student may have moved cities;
- the course is moving at university speed;
- a disappointing mark feels like a judgment on identity rather than a diagnosis of missing prerequisites.
We have not eliminated the exam.
We may have delayed it until the consequences are much more expensive.
First year as a hidden placement test
Universities describe first year as education. For many students, it also calibrates writing, mathematical foundations, self-management and help-seeking.
That calibration is noisy. First-year performance also reflects money, housing, language, disability, mental health, employment, belonging and the quality of university teaching.
A student can be academically ready and still struggle; another can enter with gaps and recover quickly. First-year marks are not a pure truth serum.
But they may be the first time many students encounter an assessment system that is both shared across a large cohort and relatively indifferent to the grades that brought them there.
The psychological invoice
Imagine being told for years that you are excellent at mathematics.
You arrive at university and receive 54.
One explanation is:
This course has exposed a specific gap in your preparation.
The explanation many students hear is:
The person you thought you were does not exist.
That is a cruel way to discover a measurement mismatch.
If the high-school signal overstated readiness, the student deserved to know earlier. If the university course is poorly taught or designed as a filter, the student deserves to know that too. If the problem is study strategy rather than subject knowledge, a placement label will miss it.
The humane response is not to make high school harsher so young people become accustomed to disappointment.
It is to make diagnosis earlier, more specific and less entangled with identity.
What would a better signal do?
I do not think the answer is one national exam that determines a life at age seventeen.
A better system would use several imperfect signals for different purposes:
- classroom grades for sustained work and growth;
- common assessments for comparability in prerequisite subjects;
- low-stakes diagnostic tasks for readiness;
- direct work samples where the capability cannot be inferred from grades;
- contextual information to avoid treating unequal opportunity as unequal potential;
- early remediation before a missing concept becomes a failed semester.
The aim is not to produce a perfect ranking.
It is to distinguish three questions we currently compress into one percentage:
What has this student achieved?
How does that achievement compare across contexts?
What support will this student need next?
One institution has already decided that identical high-school averages do not always predict identical university outcomes.
Waterloo Engineering applies what it calls an adjustment factor based partly on the historical difference between students’ high-school and first-year university averages.
That sounds like a technical admissions detail.
It is also a confession:
The currency does not have the same value everywhere.
Until the next strange question,
Osagie