The conceptual Quandary
Assessing students at Key Stage 3 presents a conceptual quandary: schools often attempt to answer several different questions with a single grade. At various points, assessment is expected to describe a student’s current performance, indicate the progress they have made over time, and predict whether they are “on track” for a future GCSE grade. These aims are not interchangeable, yet they are frequently collapsed into one system, creating confusion for teachers, students, and parents. The distinction is whether assessment should be criteria‑referenced or trajectory‑based.
A criteria‑referenced approach evaluates what a student can do in relation to clearly defined disciplinary expectations. It allows teachers to make valid statements about attainment because it is anchored in what has been learned. This model has two further category choices; is the criteria the knowledge based (facts, concepts, and procedures) or skill based (some, clear, sophisticated).
In contrast, trajectory‑based systems—often called “flight paths”—attempt to map a student’s current performance onto a predicted GCSE outcome. These systems typically rely on KS2 prior attainment and assume linear progress across five years. Grades are reduced to a ‘gradient’ indicating below, at or above expectation. There is a further category issue of whether the trajectory is compared to peers, subgroups of peers, or personal targets. While administratively convenient, they risk reducing assessment to a form of prediction rather than a reflection of learning, and they can encourage teachers to mark work according to where a student is “supposed” to be rather than what the work shows.
This can be mapped in the following way:
- Assessment for progress and attainment:
- Criteria-referenced
- Skills
- Knowledge
- Trajectory based (flight path)
- Distance from KS2 derived target
- Distance from peers
- Peers (such as class / year group)
- Peers by subgroup (such as SEN / EAL)
- Criteria-referenced
Each of these criteria will produce a different grade, whilst not all are recorded numerically. For this reason, these criteria are mutually exclusive and tell a different data story. When the data is kept in a raw state, all of the above figures can be calculated (but you do not need to track these calculations – just see that each is different to the last) in the following manner:
- Algorithms for progress and attainment:
- Criteria-referenced
- Skills = (qty of criteria met / total qty)*100
- Knowledge = (qty of criteria met / total qty)*100
- Teachers must be explicitly told the measurement; skills or knowledge
- Grades do not need to be the same as the GCSE grades
- Trajectory based (flight path)
- Distance from KS2 derived target
- Aggregate over time = (SUM(grades)/COUNT(grades))-target
- Gradient = FORECAST.LINEAR(sum of assessments, grade cells, assessment numbers)
- Final assessment only = current grade
- Teachers must be explicitly told that KS3 grades are the same as GCSE grades
- Distance from peers
- Peers (such as class / year group) = average student grade – average cohort grade
- Peers by subgroup (such as SEN / EAL) = average student grade – average cohort grade (match criteria)
- Distance from KS2 derived target
- Criteria-referenced
One grade cannot be used to indicate all of the above, because they are calculated differently, often from different ranges of data.
Is a grade a target, or a record?
Another source of conceptual difficulty is the use of target grades and cohort comparisons as measures of success. Target grades, derived from KS2 data, can create the impression that a student’s potential is fixed and measurable at age eleven. When used as the primary reference point, they shift attention away from the curriculum and towards compliance with a projected pathway. Cohort comparisons, meanwhile, can be useful for moderation and standardisation, but they are norm‑referenced rather than curriculum‑referenced: they tell us how a student performs relative to peers, not what they know or can do. Neither approach, on its own, provides a secure basis for understanding learning.
Schools therefore face a choice between several models of grading. Pure criteria‑referenced assessment is the most conceptually robust, as it aligns directly with the taught curriculum and disciplinary constructs. Mastery bands offer a simplified version of this approach, though they can drift into pseudo‑flight paths if not carefully managed. Standardised tests provide reliable comparative data but capture only a narrow slice of the curriculum. Mixed models—criteria‑based judgments moderated across the cohort—can balance validity and reliability but require strong professional agreement about standards. Flight‑path models remain the least defensible at KS3 because they prioritise prediction over learning.
How Should We Assess?
To measure attainment meaningfully, teachers must focus on what students can demonstrate now, using clear, discipline‑specific criteria. To measure progress meaningfully, schools must track how far students have moved through a well‑sequenced curriculum, comparing their current performance to their own prior performance rather than to a predicted GCSE grade. Progress is therefore a curriculum construct, not a statistical one.
When grading, departments need to make clear:
- Does this grade reflect what a student could achieve at terminal exams? – or —
- Does this grade reflect the quality of learning from the lesson? – or –
- Does this grade link the student to ex students, and what they went on to achieve based on the same standard, during this same modules?
A Long Term Peer-Comparison Model
An alternative prediction of outcome is to make a terminal projection. In this model, a curriculum needs to be quite fixed for at least five years. Case study students (identified by the various subgroups and targets we have so that a wide range of case studies is identified) should have all of their assessments kept on record along with their terminal assessment outcome (their GCSE grade). Key Stage 3 predictions can then be made in comparison to those case studies, wherein the moderation material comes directly from ex students’ assessments. This moderation material needs to have the terminal assessment of that student added to the file. When marking current students’ work, we are observing how ‘current work is comparable to the work of a student who went on to achieve grade X’. This requires a large number of assessments to be kept on record, along with a long term and stable curriculum.
This is a highly objective model, in which the teacher is removed from the subjective assessment of thinking ‘what is the current quality’ or even ‘where do I imagine this student to be in X number of years’. We are make a comparative assessment in which we identify the moderation material that the work is most similar to, and award it the grade that moderated student went on to achieve.
Final Word
These questions are not always adequately answered by schools or departments. Grading is undertaken as a begrudging work burden, but if these questions are answered properly, and the function or purpose of grades is identified, some of the barriers to marking can be lifted.