Attendance register: a record that has to stand up later

An attendance register looks like administration and behaves like evidence. Who was present decides funding in education, decides who was accounted for in a fire evacuation, and decides whether somebody can be certified as having completed training. Each of those uses demands the same thing — a record made at the time, by somebody who was there — and a register ticked afterwards from memory satisfies none of them.

What an attendance register must capture

• Who was expected, so that an absence is visible rather than simply missing from the list.
• Who was present, marked at the session rather than reconstructed later.
• The date, the session and the location, since the same group may meet several times a week.
• Who took the register, because an unsigned record cannot be relied on by anybody.
• The reason for absence where it matters — authorised or not is a different fact from present or not.
• Arrivals after the start, where lateness carries any consequence at all.

Why retrospective ticking fails

Because it is a reconstruction presented as an observation, and everybody downstream treats it as the second. The practical failures are immediate: a fire roll call against a register completed later tells you nothing about who is in the building. The slower failures are worse — funding claimed against reconstructed attendance is a finding at audit, and a certificate issued on the strength of sessions somebody may not have attended undermines every certificate you issue. The cost of doing it at the time is a minute; the cost of not doing it only appears when somebody checks.

Decide what counts as present before you need to. Half a session, joining remotely, arriving after the practical part — each needs a stated rule, because without one different people apply different standards and the register stops being comparable across groups.

Keeping it usable

1. Use a fixed set of marks rather than free text, or the register cannot be counted at the end of term.
2. Record it once, in one place — parallel registers held by two people always diverge, and nobody knows which is right.
3. Correct errors by amendment with a date rather than by overwriting, since an altered register is worth less than a corrected one.
4. Review patterns weekly rather than at the end: a pattern of absence is a signal about one person, and it is only actionable while the course is running.
5. Keep it for as long as the funding, certification or safety requirement demands, which is usually longer than anybody assumes.

The data protection part

A register is personal data, and in a school or clinic context the reasons for absence can be health data, which carries a higher bar. That means the sheet left on a desk in a corridor is a real problem rather than a pedantic one, and it means retention has to be decided rather than left to accumulate. Keeping fifteen years of registers in a cupboard because nobody decided otherwise is both a risk and, eventually, somebody's subject access request.

Ettex Sheets is enough for most registers — expected list, fixed marks, date and the person who took it — with the term's totals as a formula rather than a count by hand. Where absence reasons are recorded, they belong with the same care as any other personal data you hold.