The register on the wall works. It is the arithmetic done from it that produces arguments — and almost always for the same reason: nobody agreed what the percentage is a percentage of.
What a daily register records
One row per student, one column per day, and a status in each cell. Four statuses cover almost every school here:
- Present
- Absent
- Leave — absent, but approved
- Half day — arrived late or left early, by whatever your school’s rule is
Two more things belong in the register and are usually missing:
Who marked it. A register nobody signed is a register nobody is accountable for, and corrections become untraceable.
When it was marked. A register filled in on Friday for the whole week is not a register, it is a recollection — and it is worth being able to tell the difference later.
The monthly sheet
The document most schools actually keep: students down the side, dates across the top, one letter per cell, totals down the right.
Two things make it usable rather than merely complete:
Holidays are drawn, marked as holidays. If the 14th is blank on every row, a reader cannot tell whether the school was closed or the register was never marked. Mark it H and the sheet explains itself.
The name column stays visible when you scroll or fold. Obvious on paper, routinely broken on screen.
The percentage, and why yours is probably wrong
Three denominators are in common use, and only one of them is defensible.
Calendar days — wrong, and obviously so: it counts Sundays.
Working days — requires a model of which days your school actually runs, including a short Friday if you have one, minus declared holidays. Most schools have never written this down anywhere, and a system that invents one on your behalf will disagree with your actual calendar.
Days actually marked — the honest one. If the register was marked on 18 days and a child was present on 16, that is 89%. It is self-correcting for holidays, because nobody marks a register on a day the school is closed.
Its one weakness is worth stating plainly: a teacher who does not mark the register makes that day disappear rather than counting against the class. That is a supervision problem, and it is a better failure than the alternative, which is a child’s attendance dropping because an adult forgot.
Unknown is not zero
If nothing was ever marked for a student — a new admission, or a month nobody filled in — the honest output is no percentage, not 0%.
Printing 0% on a report card for a child whose register was never marked is a statement about that child which is not true, and it is the kind of error a parent remembers.
What a register has to survive
A student who joins in January. They should not appear on the November sheet, and their percentage should be calculated from the day they enrolled — not from the start of the year.
A student who is promoted or moves section. Last year’s register must still show them in last year’s class. If your records find a register by looking at where a child sits now, then the moment they move, their own attendance history moves with them and the old sheet empties out.
A correction. Registers get marked wrong. The question is whether a correction is possible, who may make one, and whether the fact that it was corrected survives. A school that cannot correct a register ends up with a second, informal register — which is worse.
A reasonable policy on back-dating
Corrections should be possible and bounded. Two settings cover it:
- How many days back a register may be marked or corrected — a small number, often zero or one, so that “today’s register” means today’s.
- Who may go further back than that, as a deliberate permission rather than something every teacher has.
The point is not to make correction hard. It is that a register anybody can rewrite at any time is not evidence of anything, and attendance is exactly the record schools are asked to produce as evidence.
A blank one to print
We keep a free monthly attendance register — thirty rows, thirty-one day columns, totals on the right and a key that includes holidays. Free to print and share.