Schools report reading ages to the month, which makes them sound like a measurement of something physical. They are comparisons against a sample of other children, and the precision is largely presentational.
A reading age is a position, not a quantity
A reading test produces a raw score, usually the number of items answered correctly. That number means nothing until it is placed against the scores of children of known ages.
The reading age is the age at which the average child in that comparison group produced the same raw score. It describes where a child sits in a distribution rather than what they can read.
This is why a reading age can move sharply without the child's reading changing much. A few extra items answered correctly near the middle of the distribution shifts the reported age considerably.
The norm sample sets the whole scale
Every reading test is standardised on a sample of children, and that sample defines what average means. A different sample would produce different reading ages from identical answers.
Samples age. A test standardised years ago carries the vocabulary and reading habits of that cohort, and scores drift as the population it described stops matching the children being tested.
Publishers restandardise periodically for this reason. A school that changes test provider often sees a step change in reported reading ages that reflects the new sample rather than the pupils.
Different tests measure different things
Some tests ask a child to read single words aloud. Others require silent comprehension of passages, and some measure how quickly a child reads rather than how accurately.
These skills are related but not identical, so a child can post a strong word recognition age and a weaker comprehension age. Both numbers are correct and they answer different questions.
Comparing a reading age from one test against a reading age from another therefore compares two scales that share a label. The direction of travel is more informative than the gap between them.
The error band is wider than the number implies
Any single test session contains noise: guessing, tiredness, an unfamiliar topic in a passage. Test manuals publish a confidence band precisely because one score is an estimate rather than a fact.
That band is often a year wide in either direction. A reading age reported to the month therefore carries far less resolution than its format encourages parents to assume.
Retesting the same child within a short period usually produces a different figure. The variation is a property of the instrument, not evidence that anything about the child changed.
What the number is genuinely useful for
Tracked across years with the same instrument, reading scores show trajectory, and trajectory is what teachers act on. A flat line across several testing points is a signal worth investigating.
Scores also work as a screen. A result far below the group prompts closer assessment by someone qualified to examine decoding, vision, hearing or language, rather than standing as a conclusion itself.
Treated as a verdict, the number misleads in both directions. Treated as one reading from a rough instrument, it does the job it was actually built to do.