What Cognitive Assessments Can — and Cannot — Tell You

Cognitive assessments such as the CAT4 and CogAT occupy an odd place in school life: trusted too much by some, dismissed too quickly by others. Both reactions waste what the instruments genuinely offer. Used well, a reasoning assessment adds a dimension the attainment record cannot supply; used badly, it hardens into a verdict it was never designed to be.

What they measure

Group ability tests sample reasoning — verbal, non-verbal, quantitative and spatial — under standardised conditions. Their value is that reasoning is not the same as attainment: a child can reason powerfully while attaining modestly (opportunity, language, effort and circumstance all intervene), and the reverse also occurs. Lohman (2005) argues the most informative reading is the profile: the pattern across domains, compared over time, tells a school far more than any single composite.

What they predict — and how imperfectly

Reasoning scores do correlate with later attainment: Deary, Strand, Smith and Fernandes (2007), following a very large English cohort from age 11 to GCSE, found a strong association between general ability and examination results. But strong is not total — a substantial share of the variation in outcomes lay elsewhere: teaching, effort, interest, circumstance. That gap is not a flaw in the tests; it is the space in which schools do their work.

What they cannot tell you

A score cannot say why. An unexpected result may reflect ability — or a bad morning, limited English, test unfamiliarity or disengagement. Scores carry measurement error, and extreme scores tend to drift towards the mean on retest. And a score says nothing about the qualities that research shows matter enormously for the development of talent — persistence, curiosity, productive habits — which Subotnik and colleagues (2011) place at the centre of whether ability becomes accomplishment. This is why, in our framework, no single score defines a learner and no automated suggestion closes a pathway.

Reading a score responsibly

Three practices keep cognitive data honest. Read profiles, not ranks: a spiky profile is information, not noise. Corroborate: set the score beside attainment, classroom observation and authentic work, and investigate disagreements rather than averaging them away — the disagreement is often the most useful finding. Recheck: a starting picture is a hypothesis to revise, and screening everyone rather than testing only the nominated — as Card and Giuliano’s (2016) universal-screening study showed — surfaces able learners whom referral-based systems miss.

This is how the Thinking lens works inside the Lighthouse Talent Framework: cognitive insight where an appropriate assessment is available, always read alongside the other three lenses, never alone.

References

Lohman, D. F. (2005). The role of nonverbal ability tests in identifying academically gifted students. Gifted Child Quarterly, 49(2), 111–138.
Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational achievement. Intelligence, 35(1), 13–21.
Subotnik, R. F., Olszewski-Kubilius, P., & Worrell, F. C. (2011). Rethinking giftedness and gifted education. Psychological Science in the Public Interest, 12(1), 3–54.
Card, D., & Giuliano, L. (2016). Universal screening increases the representation of low-income and minority students in gifted education. PNAS, 113(48), 13678–13683.