How can NZ schools manage NCEA change workload?

Why is the NCEA transition adding so much to teacher workload?

The workload is high because several changes are landing on top of each other rather than in sequence. The NCEA Change Programme was rephased by two years, so revised standards are being implemented progressively later this decade instead of arriving at once. Separately, the Government has confirmed NCEA will be replaced by a new senior secondary qualifications system phased in from 2028. Teachers are delivering one framework, planning for a second, and rebuilding assessment materials for both.

The day-to-day workstreams have not paused either. Departments are still working with the standalone literacy and numeracy co-requisite, redeveloped achievement standards, more digital external assessment, and greater inclusion of mātauranga Māori across subjects. Each of those means new specifications to read, fresh tasks to write, and unfamiliar criteria to mark against. The changes are individually reasonable. The problem is that they arrive in the same term.

It helps to name what is not changing. The phasing works by cohort, beginning with younger year groups, so students already working through NCEA continue on that pathway. And nothing in the transition changes what good assessment practice looks like: a clear task, a schedule that defines the evidence at each grade, consistent judgement, and feedback a student can act on. Everything below is about protecting those four things while the framework moves underneath you.

Where does the marking time actually go?

Most of it goes into internal assessment. External exams are marked elsewhere. Internals are marked by you, against a schedule you wrote, for a standard you may be teaching for the first time. The judgement itself is rarely the slow part. The slow parts are deciding what Merit actually looks like for this particular task, writing that decision down clearly enough to apply it thirty times, and then explaining it to each student individually.

Feedback is the second cost, and it is the one that quietly expands. Students need comments that are standard-referenced, so they can see why the work sits where it does and what would move it up a grade. Written from scratch for a full class, several times a term, that is evenings of work. It is also the part students cannot get anywhere else. Published exemplars and study guides show them what the standard looks like in general, but only your comment can tell them why this paragraph of theirs sat at Achieved and what specifically would lift it.

The third cost is moderation preparation. NZQA externally moderates internal assessment, so judgements have to be defensible and consistent enough to hold up when someone else looks at the same evidence. Departments that treat this as an end-of-year task pay for it twice: once in the scramble, and again in the re-marking when a sample comes back inconsistent.

How do you write a marking schedule that speeds up marking?

Write the schedule before you finalise the task, not after. If you cannot describe what Achieved, Merit and Excellence evidence looks like for this specific piece of work, the task is not finished. Working in that order surfaces the problem early, when fixing it costs an hour, rather than after thirty students have submitted, when fixing it costs a fortnight.

Make each descriptor observable. “Shows insight” is not markable. “Explains the effect of at least two techniques and links them to the writer’s purpose” is. The test is simple: could a colleague who has never taught this class apply your descriptor and land on the same grade? If not, keep rewriting it. That is the same discipline that underpins expert essay marking anywhere, and it is what makes a schedule reusable next year rather than disposable.

Then mark two or three scripts against the draft schedule before you touch the rest of the pile. Almost every schedule needs an adjustment after contact with real student work, and that adjustment is far cheaper before you have committed to thirty judgements you would otherwise have to revisit.

• Draft the schedule before you finalise the task
• Phrase each grade descriptor as something you can point to in the script
• Mark two or three scripts first, then adjust the schedule
• Keep one annotated exemplar at each of Achieved, Merit and Excellence
• Store the schedule and exemplars where the whole department can find them

How do you keep judgements consistent across a department?

Consistency comes from shared reference points, not from shared intentions. Two teachers who agree in a meeting will still drift apart by week six. What stops the drift is a small set of annotated exemplars, one per grade, that everyone marks against and can point to when a judgement is questioned.

Run a short calibration early in the term rather than a long one at the end. Take four scripts, have each marker grade them independently, then compare. Where you disagree, you have found the sentence in the schedule that is doing the least work. Half an hour spent there saves days of re-marking later, and it is about the closest thing to a free improvement available in a busy term.

Keep the trail as you go. Note which part of the evidence justified each grade at the point of marking, while the reasoning is fresh. That turns moderation preparation into a collection job rather than a reconstruction job, which matters most when the standard is new and your interpretation of it is still settling.

How do you make feedback faster without making it thinner?

Build a comment bank tied to the schedule, not to the task. A comment phrased against a grade descriptor, such as “you have identified the technique, and to reach Merit you need to explain its effect on the reader”, transfers to every assessment that uses that standard. A comment phrased against one specific text dies with the task.

Aim for one strength, one next step, and where you can, a sentence pointing at the evidence in the student’s own work. That last part is what makes feedback land. Past about three points students act on none of them, so the shorter, sharper comment is usually the better one as well as the faster one.

Then reuse deliberately. A department that maintains a shared bank for each standard it teaches is not cutting corners. It is refusing to write the same sentence thirty times. The time that frees up goes back into the borderline judgements and into the students who need a conversation rather than a comment.

How does a marking assistant fit into this?

Everything above works with paper, a shared drive and a department meeting. Software helps only after that thinking is done, by taking over the repetitive part of applying it. Since this site is run by the team behind one such tool, it is worth being specific about what Jeddle’s marking assistant, JeddAI, does and does not do, so you can judge whether it fits your department.

You set up your own success criteria and comment bank for a standard, mirroring the Achieved, Merit and Excellence descriptors you have already written. JeddAI then drafts feedback and a suggested judgement against that reference for each piece of student work, and you review, edit and confirm every result. It does not mark autonomously, and it does not replace NZQA moderation. What changes is the starting point: you are editing a draft that already cites your criteria instead of composing from a blank page.

A sensible trial is one internal assessment and one class. Mark that class yourself as well on the first run and compare the two, which is the same calibration exercise you would run with a colleague. If the drafts sit where your judgements sit, share the criteria across the department and reuse them for the next standard. If they do not, it is usually the schedule that needs tightening, and that is worth knowing either way.