Moving from Numerical Marking to Grading: A Better Evaluation System

For decades, the numerical marking system has been the default way universities evaluate student performance. A student scores 73 marks, another gets 76 – and suddenly, those three points become the difference between one academic fate and another. But how meaningful is that difference, really? Across campuses worldwide, educators and researchers are increasingly questioning whether numerical marks give an accurate, fair picture of what students actually know and can do. The shift toward letter-based grading systems is gaining ground – and there are strong reasons why universities should take it seriously.

Problems with the numerical marking system 🔗

Numerical marking has served higher education for generations. It feels precise, measurable, and objective. But that sense of precision is, in many ways, an illusion.

Inconsistency across evaluators 🔗

A core problem is inter-rater inconsistency. The same piece of student work can receive different scores depending on who is marking it. Durham University’s marking guidelines acknowledge that marking is fundamentally “a matter of qualitative academic judgment” – yet the numbers used to express that judgment create a false appearance of scientific precision. The number 70, for instance, does not mean a student met 70% of the learning outcomes; it is simply a conventional marker for a threshold of quality.

This becomes a practical problem when multiple instructors evaluate students across sections, semesters, or departments. Research published in Assessment and Evaluation in Higher Education found that numerical scoring relies on theoretical assumptions that are rarely tested, and that aggregating marks from multiple assessment components introduces compounding errors in final scores. In other words, the more components a final score draws from, the less reliable it becomes.

The fairness problem 🔗

Beyond inconsistency, numerical marks often fail students whose strengths lie outside the narrow formats exams test. A student who performs poorly on written assessments but excels in applied, practical tasks may receive a score that drastically underrepresents their actual competence. The AACSB has noted that both inconsistency across courses and grade inflation have affected universities in many parts of the world, with the core difficulty being that absolute grading is an elusive goal – without controlling for exam difficulty, marks mostly reflect relative performance anyway.

A further dimension of unfairness comes from how tightly clustered numerical scores can distort outcomes. As Durham’s analysis shows, if strong students’ marks cluster in a narrow band while weaker marks spread widely, even consistently high-performing students can be unfairly pulled down by a single moderate score. The mathematics of aggregation works against them – not because of their ability, but because of how the scale is structured.

What marks don’t measure 🔗

Numerical marks also tend to capture only a limited slice of a student’s abilities – primarily recall, comprehension, and problem-solving under exam conditions. According to research published in Assessment & Evaluation in Higher Education, the origins of grading were partly to enable communication between teachers about individual student progress – a function that a raw percentage often handles poorly. Qualities like creativity, collaboration, analytical thinking, and communication are routinely excluded from the numerical score, leaving the mark as only a partial reflection of a student’s learning.

Advantages of grading: reducing errors and improving accuracy 🔗

Moving from numerical scores to a letter-based grading system addresses many of these weaknesses directly. Rather than forcing evaluators to distinguish between a 74 and a 76, grades group performance into broader, more defensible categories – typically A, B, C, D, and F – each anchored to clear performance descriptors.

Fewer subjective errors 🔗

One of grading’s most significant practical benefits is that it absorbs small evaluative differences that are essentially meaningless. An evaluator deciding between a B and a B+ is making a less consequential judgment than one deciding between 71 and 73 marks. The AACSB points out that letter grades are themselves an acknowledgment that assessment instruments are prone to measurement error – and that trying to report performance more finely than a letter scale allows only compounds that error.

A more holistic view of student performance 🔗

Grading systems are typically built around broader criteria that encompass a range of assessment types – exams, assignments, presentations, projects, and participation. This gives a more complete picture of a student’s capabilities than any single number can. The Association for Career and Technical Education Development (ACAD) describes effective grading as “bias-resistant” – grounded in valid evidence of a student’s actual content knowledge rather than factors that reflect implicit bias or a student’s environment.

Standardization across institutions 🔗

Letter grades also carry a shared meaning that travels across departments, universities, and even national borders. The Week notes that grading systems are universal in nature, and using a system understood across institutions makes it easier for students and employers to interpret academic achievement meaningfully. This is especially relevant as higher education becomes increasingly international.

How grading works: from marks to letters 🔗

The practical mechanics of grading are straightforward. Universities typically convert raw numerical marks into letter grades by assigning each letter a score range. The most widely used conversion structure looks like this:

• A (Excellent): 90-100%
• B (Good): 75-89%
• C (Average): 60-74%
• D (Below Average/Pass): 50-59%
• F (Fail): Below 50%

Some institutions add plus and minus designations (A+, A−, B+, etc.) for additional granularity, though the core categories remain consistent.

The GPA system 🔗

Many universities take grading a step further through the Grade Point Average (GPA) system. Each letter grade is assigned a numerical point value – typically A = 4.0, B = 3.0, C = 2.0, D = 1.0, F = 0 – and the GPA is calculated by averaging weighted values across all courses taken. International credential evaluators like IEE use this system to translate performance across different national grading systems into a common standard, which is essential for student mobility and graduate admissions.

The European Credit Transfer and Accumulation System (ECTS), adopted by universities across Europe, uses a similar letter-grade structure (A through F), enabling student records to be understood uniformly across member countries. Norway, for example, transitioned from a numerical 1.0-6.0 scale to the ECTS letter system in the early 2000s, precisely to improve cross-institutional comparability.

Absolute vs. relative grading 🔗

Universities can implement grading in two broad ways. Absolute grading assigns letter grades based on fixed thresholds – a student scoring above 75% receives a B regardless of how peers performed. Relative grading (sometimes called norm-referenced or curve grading) assigns grades based on a student’s standing within the class distribution. Most higher education systems favor absolute grading as it assesses a student’s mastery of content on its own merits, rather than penalizing students in high-achieving cohorts or rewarding them in weaker ones.

Implementation strategies: how universities can make the shift 🔗

Transitioning from a numerical marking system to a grading framework is not simply a technical change – it requires deliberate institutional planning, faculty alignment, and clear communication with students.

Step 1: Establish clear grading criteria 🔗

Before any transition begins, universities need to define what each letter grade actually means in terms of learning outcomes. Vague categories invite the same subjectivity that plagued the old system. Professor Thomas Guskey, writing in Phi Delta Kappan, recommends that institutions adopt a grading scale with four to seven performance categories and build those categories around specific, articulated standards. A clear purpose statement – explaining what a grade represents and what it does not – is the essential foundation.

Step 2: Secure faculty buy-in through training 🔗

Faculty are central to any grading reform. Northern Illinois University’s Center for Innovative Teaching and Learning emphasizes that grading policies can inadvertently perpetuate achievement disparities even when faculty believe they are being fair. Structured workshops, calibration exercises, and shared rubrics help ensure that instructors across departments apply grade criteria consistently. The U.S. Department of Education similarly stresses that the key to meaningful grading reform lies in scale – educators across departments and universities need to collaborate to establish consistent, meaningful standards rather than act individually.

Step 3: Pilot and phase the transition 🔗

An abrupt system-wide switchover creates confusion and resistance. A phased rollout – starting with select departments or course types before expanding institution-wide – allows problems to be identified and corrected early. Edutopia’s reporting on school districts that have made this shift notes that grading reform, when done gradually and with teacher involvement, shifts not just the grading system but the entire culture of assessment – moving educators toward analyzing student data and closing learning gaps rather than simply assigning numbers.

Step 4: Communicate clearly with students 🔗

Students need to understand how grades will be calculated, what each grade reflects, and how changes affect their academic records. Transparency reduces anxiety and builds trust. Columbia University’s Center for Teaching and Learning recommends sharing detailed rubrics with students upfront, reducing grade-related stress, and shifting student focus from score maximization to genuine learning. This kind of transparency also transforms grade conversations: rather than asking how to recover lost points, students begin asking what they need to understand better.

Step 5: Monitor, evaluate, and adjust 🔗

Implementation does not end at rollout. Universities should track grade distributions across departments, gather student and faculty feedback, and compare outcomes over time. Monitoring helps institutions catch grade inflation early – a concern that even elite universities are grappling with. Harvard’s Faculty of Arts and Sciences, for instance, is currently debating a proposal to cap top grades in response to evidence that over 60% of undergraduate grades had become A’s, effectively rendering the highest distinction meaningless. Grading reform is not a one-time event; it requires ongoing institutional attention.

Why this shift matters for higher education 🔗

The move from numerical marking to a grading system is ultimately about making evaluation more honest and more useful – both for students trying to understand their progress and for institutions trying to communicate that progress to the world. Grades, when designed thoughtfully, reduce arbitrary scoring differences, absorb measurement error, reflect a broader range of student capabilities, and travel across institutional and national contexts in a way that raw percentages rarely do.

That does not mean grading is without challenges. Grade inflation, inconsistent application, and potential overemphasis on letter outcomes are real risks. But these are problems of implementation, not of the grading concept itself. With clearly defined criteria, well-trained faculty, phased adoption, and transparent communication with students, universities can build evaluation systems that are fairer, more accurate, and more meaningful than the numerical approach they replace.

What do you think? If two students score 71% and 74% respectively, should those three marks genuinely determine different academic outcomes – or does collapsing both into a B-grade serve them better? And how should universities balance the push for standardized grading with the need to preserve meaningful academic distinctions?

References

1. https://www.durham.ac.uk/departments/academic/common-awards/policies-processes/assessment/marking-numerical/
2. https://researchers.mq.edu.au/en/publications/using-marks-to-assess-student-performance-some-problems-and-alter
3. https://www.aacsb.edu/insights/articles/2021/10/grade-debate-absolute-and-relative-measures
4. https://www.tandfonline.com/doi/full/10.1080/02602938.2022.2134552
5. https://acad.org/resource/the-time-is-now-for-equitable-grading-in-higher-education/
6. https://theweek.com/education/1022248/pros-and-cons-of-the-letter-grading-system
7. https://iee.com/blog/gpa-and-global-grading-scales-explained/
8. https://en.wikipedia.org/wiki/Grading_systems_by_country
9. https://kappanonline.org/addressing-inconsistencies-in-grading-practices/
10. https://citl.news.niu.edu/2021/10/26/fair-consistent-transparent-grading/
11. https://www.ed.gov/about/homeroom-blog/addressing-grade-inflation-collective-action-problem
12. https://www.edutopia.org/article/transitioning-to-evidence-based-grading/
13. https://ctl.columbia.edu/resources-and-technology/resources/grading-for-learning/
14. https://news.harvard.edu/gazette/story/2026/03/plan-to-rein-in-inflated-grading-explained/