Quick answer

Record both dates, the exact result, the format, the purpose and any convention applied. Five fields make the calculation auditable.

What a complete record contains

Five elements make an age calculation reproducible. The date of birth and its source. The reference date and why that date governs. The exact result. The format or convention applied, if any. And the purpose — the instrument, policy or decision the age was calculated for.

Any reviewer with those five can rebuild the number without contacting you. Missing any one of them and they are guessing. The most commonly omitted is the reference date, which is also the one that makes the record decay silently as time passes.

Use this guide with the right tool: Open the Chronological Age Calculator for chronological age on an assessment date. If the question shifts, compare it with the Age on a Date Calculator or the Test Age Calculator.

Screenshots are not documentation

A screenshot of a calculator result shows the output and often not the inputs. It cannot be searched, cannot be re-run, and cannot be checked against a corrected date of birth if one later emerges.

Record the values as data. Where you want a durable pointer to the exact calculation, the shareable result link encodes the inputs in the address, so opening it reproduces the calculation rather than an image of it. Keep the link alongside the recorded values, not instead of them.

A blank notebook, pen and filing folder on a desk
A complete record lets anyone rebuild the calculation — inputs, rule, and result.

Note the unusual, not the routine

Documentation that records everything at equal weight is hard to read. What deserves an explicit note is anything a reviewer would not assume: an estimated date of birth, a multi-session administration, a nonstandard reference date, an applied rounding rule, a boundary case, or a discrepancy between two source records.

Routine calculations need only the five fields. It is the exceptions that generate the questions, and a one-line note at the time costs far less than a reconstruction eighteen months later.

Corrections and version history

When a date of birth is corrected, every age derived from it becomes wrong. A record that lists the source of each date makes that a tractable problem — you can find the affected calculations. A record that lists only results does not.

Where a figure is revised, keep the original alongside the correction with dates for both, rather than overwriting. An audit trail that shows a value changed and why is stronger than one that appears to have always been right.

Worked example

A complete entry reads: DOB 12 February 2018 (from birth certificate on file); assessment date 16 August 2026 (first session, per manual §2.3); chronological age 8;06;04; exact calendar method, no rounding applied; calculated for norm-table selection. Any reviewer can rebuild that from the record alone.

The complete entry pairs a birth date of 12 February 2018 with an assessment date of 16 August 2026 and records 8;06;04 by the exact calendar method with no rounding, for norm table selection. Rebuild your own record with its sources, method and purpose stated just as plainly.

The same record, rebuilt after a correction

Take the worked example's complete entry — date of birth 12 February 2018, assessment date 16 August 2026, result 8;06;04 — and suppose the birth certificate is re-read and the date was 21 February 2018 all along: a transposition of the day digits, the commonest kind of copying error. The rebuilt entry keeps four of its five fields and changes only the first, yet the result is 8;05;26. The correction is nine days, but it does not read as nine days anywhere in the figure: it has moved the months field down and the days field up, so a reviewer comparing the two results sees what looks like a different calculation rather than a small amendment to one. This is the practical reason the version history matters. Any norm table chosen from 8;06;04 has to be re-examined against 8;05;26, and only a record that kept the original entry, the correction and the date of the correction can show which decisions were taken on which figure.

The reference date is the second fragile field, and it fails without anything being wrong. Assess the same child — the original 12 February 2018 date of birth — in a second session on 2 September 2026 instead of 16 August, and the exact age is 8;06;21. Seventeen days separate the sessions. If a norm boundary sat at 8;06;15, the two sessions would place the same child on opposite sides of it, with both calculations correct. A file that records only one age, without saying which session's date produced it, cannot explain the discrepancy when it surfaces; a file that records the governing date alongside each result explains it in one line. Multi-session administrations are on this guide's list of things worth a note for exactly this reason: the unusual thing is not the arithmetic, it is that there were two dates the arithmetic could honestly have used.

Neither case required a new method. The five fields absorbed both: the corrected date of birth with its source, the reference date with its reason, the exact result, the convention, the purpose. What the cases add is timing — the fields earn their keep not when the calculation is made but months later, when someone who was not in the room has to rebuild it and find out whether the number in front of them is the number that was decided on.

The purpose field, the one most often treated as decoration, is what stops a figure migrating. The corrected record — 8;05;26 on 16 August 2026 — answers a question about an assessment, calculated for norm-table selection. Put the same corrected date of birth to a different purpose, an age-group check against a 1 January 2026 cutoff, and the governing date changes with the purpose: the age is 7;10;11, a figure that appears nowhere in the assessment file and should not. A result lifted from one record into another context carries its reference date invisibly or not at all; the purpose is the field that tells the next reader which question the number was the answer to.

Common mistakes to avoid

  • Recording the result without the reference date. The age is only true on one date. Without it the record cannot be verified and quietly goes stale.
  • Storing a screenshot instead of the values. An image cannot be searched, re-run or corrected. Record the data and keep any link as a supplement.
  • Overwriting a corrected figure. Keep the original and the correction with dates for both. A visible trail is stronger than a clean one.

Use the right calculator

Chronological Age Calculator gives chronological age on an assessment date, Age on a Date Calculator gives it on a past or future date, and Test Age Calculator gives calendar age when a test is administered.

Chronological Age Calculator
Clinical & AcademicChronological Age Calculatorchronological age on an assessment date
Age on a Date Calculator
Universal ChronologyAge on a Date Calculatorchronological age on a past or future date
Test Age Calculator
Clinical & AcademicTest Age Calculatorcalendar age at the date a test is administered
Stays in the tab: the birth and assessment dates you enter for a record are calculated in your browser, with no account required and nothing sent to a server.

Frequently asked questions

How much detail is too much?

Record the five core fields always, and add a note only for anything a reviewer would not assume. Uniform verbosity makes files harder to use.

Should I record the calculation method?

Name it if it is anything other than exact calendar age, and always if a rounding or padding convention was applied.

What if two records give different dates of birth?

Record both, note the discrepancy and its sources, and use whichever the governing authority accepts. Do not silently pick one.