Two children can score exactly the same on a test and need completely different help. Here's how a genuinely adaptive system tells them apart.
25 August 2026 · 5 min read
Picture two Year 5 children who each get 6 out of 10 on the same fractions test. Same score, same worksheet. On paper, they look identical.
One of them made a careless slip on two easy questions and ran out of time on two harder ones — she understands fractions fine. The other genuinely doesn't understand how to combine fractions with different denominators, and got the easier questions right mostly by luck.
A score can't tell those two children apart. Finding out what actually happened — and doing something different for each of them — is the whole point of an adaptive system. We touched on why a score alone isn't enough in what is adaptive learning in maths; this is about how the "finding out" part actually works.
"Fractions" isn't really one skill — it's dozens of smaller ones stacked on top of each other: recognising a fraction, finding an equivalent fraction, comparing fractions, adding fractions with the same denominator, adding fractions with different denominators, and so on. Each of those is a distinct thing a child either has or doesn't have yet.
A useful gap-finding system tracks each of those smaller skills separately, rather than lumping them all under one "fractions" score. That's what lets it say something precise — not "your child needs more fractions practice," but "your child is confident with equivalent fractions but hasn't yet secured adding fractions with different denominators." One of those is actionable. The other isn't.
Not every wrong answer is equally informative, but many of them are. Take a child who calculates 1/2 + 1/3 as 2/5. That's not a random slip — it's a specific, well-documented misconception: treating the numerators and denominators as two separate whole numbers to add, rather than finding a common denominator first.
A system that recognises this pattern doesn't need to guess. It can flag the likely misconception directly from the answer itself, the same way an experienced tutor would glance at a child's working and immediately know what's going on.
The other half of telling those two children apart is knowing the difference between a one-off and a trend. Everyone — adults included — makes the occasional careless error under time pressure. That's not a gap, and treating it like one just wastes a child's time re-covering something they already know.
A gap shows up as a pattern: the same type of mistake recurring across different questions, different sessions, sometimes weeks apart. That repetition is the signal. It's why a genuinely adaptive system looks at a learner's history, not just their last answer, before deciding something needs attention.
There's a detail that's easy to miss: getting a concept right once doesn't mean it's secure. A child can answer a question correctly today and quietly lose that understanding over the following weeks if it isn't revisited — which is exactly why children so often relearn the same thing every September.
A good adaptive system doesn't just close the book on a skill the moment a child gets it right. It checks back on it later, at increasing intervals, to confirm the understanding has actually stuck — rather than assuming a single correct answer means the job is done.
If you're trying to work out what your own child actually needs, the same logic applies without any software: don't just look at the score, look at which questions were wrong and whether there's a pattern to them. One wrong answer on a topic they normally do well in is probably nothing. The same kind of mistake showing up three times in a row is worth a proper look.
Grasp Maths tracks each skill separately rather than scoring by topic, checks wrong answers against known misconception patterns, and distinguishes a one-off slip from a genuine gap by looking at the pattern over time — not just the last question. Where a gap is confirmed, that's when scaffolding and a step back to the prerequisite skill actually help, which we covered in why "drill and kill" maths apps get it wrong.
Don't just practise more. Practise what matters next.
Try a free Grasp Maths practice session and see how the engine builds a skill-by-skill picture, not just a score.