What K1 Assessment Scores Actually Predict About Primary Performance

Over twelve years, I informally tracked outcomes. I talked to teachers at our feeder primaries. I paid attention to what happened to the children who had passed through my assessment room. And I built up, not a dataset, not rigorous research, but a pattern of observations that challenged several things I had believed when I started.

The core question I was trying to answer was: what, in my assessment of a three-year-old, actually predicted how she would perform in primary school?

The answer was not what I had been trained to look for.

What did not predict primary performance as strongly as expected

The specific skills assessed. Counting, colour identification, shape recognition, vocabulary on demand — these had a weaker predictive relationship to primary performance than the educational theory behind them implied. A child who couldn't count to 20 at three could count to 20 at five, and then to 100, and then do arithmetic. The pre-academic content I was testing was not a reliable signal of future academic capacity.

The school the child attended. This was the most uncomfortable finding. Children who attended our school — a well-resourced Band 1 institution with experienced teachers and good provision — did not, as a group, outperform comparably-assessed children who attended less prestigious schools, when I could track those comparisons. The school effect was present but much smaller than the cultural narrative around school selection implied.

What did predict primary performance

The quality of the child's engagement with novel tasks. Not whether she succeeded, but how she approached something she hadn't seen before. Did she attempt it? Did she stay with it? Did she recover from error and try again, or did she shut down? This was visible in three-year-olds and it predicted, with reasonable consistency, how they would handle primary school.

The warmth and groundedness of the child's relationship with the attending parent. Assessable from the waiting room, from the goodbye moment, from the child's reference to the parent during the assessment. Children with secure, warm parental relationships were more willing to engage with the unknown. They had a safe base to explore from.

The child's language — not vocabulary size but quality of engagement with language. Did she use language to explore ideas, to ask genuine questions, to describe her own experience? Or did she produce language only in response to direct prompting? The former predicted genuine academic engagement; the latter predicted test-taking skill without depth.

The finding that most surprised me

Some of the children I had most doubted — who had presented immaturely, who hadn't performed well in the structured components — turned out, by P3 or P4, to be among the stronger students at their schools. The developmental lag had resolved. The underlying capacity that the three-year-old assessment had failed to reach had emerged.

And some of the most impressive three-year-olds — the ones who had been through intensive preparation and performed brilliantly — hit walls at P4 or P5 when the work became genuinely conceptually demanding and the preparation-based advantage ran out.

The pattern underneath the pattern

What I was consistently seeing, across years, was that the family was the primary predictor. Not the school. The family.

The children with curious, engaged parents who read and thought and had genuine conversations — these children did well regardless of what school they attended.

The children from families where the primary investment was in credentials, performance, and external assessment rather than genuine intellectual engagement — these children often plateaued when the external scaffolding was removed.

I cannot say this in a way that doesn't sound like a comfortable middle-class truism. "Be an engaged, curious parent and your child will be fine." I know this lands as self-evident. But I saw the evidence of it walk through my door for twelve years, and the self-evident turned out to be correct.

The school you get your child into matters less than you believe. The family you are matters more.