Evidencing Attendance and Progress in UAE Schools

Every school can produce evidence given a fortnight. Someone exports three spreadsheets, reconciles them by hand, and assembles something defensible. The question is what you can produce on the day you are asked — and whether it holds up when someone reads it closely.

The distinction between maintained evidence and reconstructed evidence is the whole of it, and it is worth being precise about why one is stronger.

Why reconstructed evidence is weaker

Reconstruction produces the right numbers. What it cannot produce is the timeline. A maintained record shows that an attendance pattern was visible in March; a reconstructed one shows only that the pattern existed. Those are different claims, and only the first demonstrates that the school was in a position to act.

The same applies to a support plan with review dates recorded as they fell due, versus a plan whose reviews are documented in a single sitting afterwards. Both may describe the same provision. Only one shows a cycle that actually ran.

This is why “inspection-ready” is not a state a school enters shortly before a visit. It is a property of records that were kept properly all along, and it is the reason the work is continuous rather than seasonal.

Attendance: what has to be defensible

Four things, and most systems handle only the first two well:

• Session-level marking, not just a morning register. A single daily mark cannot distinguish a student who was present all day from one who was on site but not in lessons. Where a school marks against the timetable, internal absence becomes visible.
• Reasons captured at the point of marking. Authorised, unauthorised, medical, leave — recorded when the mark is taken, not reconstructed at the end of term when nobody remembers.
• Unmarked sessions visible as unmarked. This is the most damaging failure mode there is. A system that treats a forgotten register as “all present” produces data nobody should trust, and the error is invisible.
• Corrections with a trail. Marks get corrected; that is normal. What matters is that the change records who made it and why, and that the original remains visible.

There is also a calculation question that trips schools up: a student who joined in March should have their percentage calculated from their start date. Calculating from September makes a mid-year joiner look like a serious attendance concern, which in a market with high in-year mobility misrepresents a meaningful share of the roll.

Progress: two measures, not one

A grade on its own answers “where is this student now”. It says nothing about distance travelled, and distance travelled is usually the more revealing of the two.

Holding attainment against expected standard separately from progress from a baseline lets a school make both arguments — including the one about the student who arrived two years behind and gained three, which a single grade actively conceals.

This requires:

• A baseline recorded on entry, including for mid-year joiners;
• Historic attainment attached to the student, not the class, so it survives year-group changes;
• Cohort history retained long enough for year-on-year comparison to mean something;
• Prior attainment from another school recordable, so progress is measured from a real starting point rather than a blank field.

How assessment is structured covers the mechanics of keeping these separate.

The connections are where the argument lives

Attendance and attainment are individually unremarkable. Read together they answer questions neither can answer alone: whether the students whose attainment dipped were also the students whose attendance did, whether an intervention coincided with a change, whether a behaviour pattern tracks a subject or a time of day.

Those connections only exist if the data sits on the same student record. Where attendance lives in one system and assessment in another, the correlation requires someone to suspect it first and then go looking — which means the ones nobody suspected are never found.

What to keep continuously

1. Session-level attendance with reasons and a correction trail.
2. Assessment history attached to students, retained across year groups.
3. Baselines on entry, including mid-year joiners.
4. Support plans with provision delivered and reviews as they fall due — not written up afterwards.
5. Behaviour and wellbeing entries, positive as well as negative.
6. Access logs, so who viewed or changed a record is answerable.
7. Documents generated from the record, so what left the school matches what the school holds.

The practical test

Ask for something specific and see how long it takes: attendance and attainment for one year group, compared with the same year group two years ago, broken down by section.

If that is a report someone runs, the records are maintained. If it is a project someone schedules, they are reconstructed — and the fortnight it takes is not the real cost. The real cost is that the same question, asked during a meeting, goes unanswered.

Reporting and analytics is built around that distinction, and the questions worth asking before choosing a system cover how to test for it before committing.