The attendance register produces one figure at the end of every month — 92% in July, 88% in August — and that figure travels: into the newsletter, the management meeting, the return the school submits upward. It is also the least informative number the school owns.
A percentage that pools a Class 3 pupil who missed two days with fever and a Class 9 student who has missed every Monday this term describes neither of them. Reading attendance properly means replacing one figure with a small set of questions: which group of students, over what period, counted how, and what should happen when the answer changes.
Check the denominator before you trust the percentage
Two schools can both report 90% and mean entirely different things. Before comparing months, sections or years, confirm four things about the number in front of you.
- Who is included. Does the total count students admitted mid-month, students who have left but are still on the rolls, students on long approved leave? Each choice moves the denominator.
- How a day is counted. Is a half-day present, absent, or half a day? Is a late arrival present? Whatever the rule, it has to be the same rule in every classroom register and in the office summary.
- Which register is the source. If teachers mark on paper and the office retypes it later, you have two datasets and one of them is a transcription. Reports built on the second inherit every typing error.
- When it was generated. An export taken on the 28th and one taken on the 30th are not the same month. Note the cut-off date on the report itself.
Move from one school-wide figure to cohorts
The school average is a lagging summary. The useful view is the spread underneath it. Slice the same data by groupings your timetable already uses.
- Class and section. A section sitting consistently below its peers usually has a specific cause — a period it misses most, a transport route, or a teacher on long leave.
- New admissions against continuing students. A child joining mid-term is navigating an unfamiliar building, timetable and friendship group. Their first weeks deserve their own line on the report.
- Day scholars against hostel residents and regular bus users. A route running late, a changed pickup time, or an illness cluster in a hostel shows up here first.
- Students with approved leave patterns — family travel, medical leave, sports or competition duty — so that expected absence is not read as disengagement.
Slice by these groups to find questions, not to label children. An attendance report is a prompt for a conversation with a class teacher, never a verdict on a family.
Separate chronic absence from occasional absence
Occasional absence and chronic absence look identical in a monthly average and need completely different responses.
Occasional absence is usually explainable and self-correcting: illness, a festival, a hospital visit. Chronic absence is a pattern — repeated short gaps, the same weekday, or absence that climbs as the term progresses. The first needs a record; the second needs a person.
The practical test is the shape of several months, not one. A student who missed four days in August and four in September is a different case from one who missed eight days in August and none since. Plotted across the term, the difference is obvious.
Every school should agree its own line for when absence becomes a case worth following up, write it down, and apply it the same way to every child. A threshold borrowed from elsewhere, or applied only to the students someone already worries about, produces both unfairness and blind spots.
Look at absence by subject and period, not only by day
A day-level percentage cannot see the most fixable problem in a school timetable: a student who misses the same weekly period every time. Marked absent once a week, they are unremarkable against a school-wide figure, but it is the same lesson every week — the one they never receive.
Look for clusters at:
- First period, where transport delays and late arrival concentrate.
- Last period, where early pickups and outside coaching schedules bite.
- Double practicals and lab sessions, where a missed session is hard to make up.
- The days before and after unit tests, where absence distorts both the assessment and the child’s confidence.
Period-level attendance is where an attendance report stops being a compliance document and starts being a timetable review.
Put attendance beside assessment data
Attendance and marks are usually kept in separate places and discussed in separate meetings, which is exactly backwards. A student who was absent the week a topic was introduced and is now failing that topic may not be a student who cannot do the work.
Read together, the two registers produce sentences you can act on: these children missed the week we taught fractions; this batch missed the revision sessions before the unit test. Read separately, they produce adjectives about effort.
Set thresholds that trigger a defined response
A threshold is only worth setting if something happens when it is crossed, and someone knows it is their job. Before configuring an alert, write the response down.
- Single unexplained absence. The class teacher confirms the register at the end of the day, in case the mark itself is wrong.
- A pattern across several consecutive days. The class teacher or coordinator contacts home, records the reason, and closes the case.
- Sustained or recurring absence. A support conversation with the family, and a check on whether something in the school day — a period, a route, a peer problem — is part of the cause.
- Whole-cohort absence. If a whole section is below its peers, the response is an operational review, not a list of individual calls.
Each needs an owner, a review window and a closing state. Without a closing state, the dashboard fills with warnings nobody clears and staff stop trusting it.
Give the review a rhythm
Weekly, work from an exception list: new cases, cases in progress, cases resolved, and anything that has sat untouched beyond its window. It should take minutes, not a meeting.
Monthly, step back to cohorts: which sections, routes and year groups moved, and whether the movement is seasonal, structural or a data-quality artefact.
At the end of the term, review the review: how many alerts led to a real conversation, how many were false alarms caused by a bad mark, and which thresholds need tightening or loosening. Safety concerns should never wait for this rhythm — they escalate immediately, outside the attendance report.
Data traps that quietly corrupt attendance reports
- Duplicate student records from an admission entered twice, so one child appears as one present and one absent.
- Students marked in two registers — the classroom’s and a hostel or activity register — and reconciled by hand, or not at all.
- Bulk marking at the end of the day, which loses the specific period a student actually missed.
- Corrections applied to last month’s register after the report was generated, so the report and the register no longer agree.
A monthly reconciliation of one section against its paper register catches most of these.
A short checklist before your next attendance review
- The definition of a present day is written down and applied everywhere.
- The report states its date range and cut-off.
- Absence is shown by class, section and period, not only school-wide.
- Chronic absence is distinguished from occasional absence by pattern, not by one month’s total.
- Attendance sits in the same review as assessment data.
- Every threshold has an owner, a response and a closing state.
- Safety concerns bypass the monthly cycle.
- The numbers going into any external return match the register they came from.
Scholva’s analytics module recomputes attendance daily across classes, sections and periods, so the exception list a coordinator opens on Monday is built from the same registers teachers marked in, rather than from a spreadsheet compiled over the weekend.
Related reading: The School Attendance Workflow: From Classroom to Parent Alert and Early-Warning Signals School Leaders Should Track.