What concept-based learning is
Most school learning is organised around topics: chapters in a book, units in a syllabus, sections of a test. Concept-based learning changes the unit. It organises teaching, practice and — crucially — measurement around concepts: single, transferable ideas that questions are built from.
Definition
Concept-based learning is an approach in which what a student is taught, practises and is assessed on is organised around individual concepts — discrete, transferable ideas — rather than around chapters, topics or overall scores. Because concepts depend on other concepts, it also allows a difficulty to be traced back to the earlier idea causing it.
The difference sounds small. It is not. Consider one topic from a class 8 maths syllabus:
| Topic | Some of the concepts inside it |
|---|---|
| Profit, Loss and Discount | What a percentage represents · Finding a percentage of a quantity · Fraction–percentage equivalence · Cost price and selling price · Discount as a share of the marked price · Profit or loss as a percentage of what · Working backwards from a final price |
A student scores 40% on the chapter test. What do you now know? That they are weak at profit and loss — which is a label, not a diagnosis. They might be entirely secure on six of those seven concepts and stuck on one. They might be fine on all seven and unable to hold a two-step problem together. Or the real problem might not be in this list at all: it might be fractions, from three years ago.
A topic-level score can only ever tell you where a student failed. It cannot tell you what to do about it.
Concepts depend on other concepts
The second half of concept-based learning is the part that makes it powerful: concepts are not a flat list. They form a dependency map, where each idea rests on earlier ones.
- FractionsA part of a whole, and how parts combine.
- PercentageDepends on fractions — a percentage is a fraction with a fixed denominator.
- Profit, loss and discountDepends on percentage, plus knowing which quantity the percentage is taken of.
- Multi-step word problemsDepends on all of the above, plus holding several steps together at once.
Once the map exists, a piece of reasoning becomes possible that a topic-based system simply cannot perform. If a student is unreliable on multi-step word problems, you can ask: are they reliable on the concepts underneath? If percentage is also unreliable, the word problems are not the issue. If fractions are unreliable too, then neither percentage nor word problems are the issue — and every hour spent practising them is an hour spent building on sand.
That downward trace is how a foundational learning gap is found.
Concept-based learning vs. covering the syllabus
These are not in competition. The syllabus still says what has to be taught and by when. Concept-based learning changes the resolution at which understanding is tracked underneath it — which means a class can keep moving through the syllabus while individual weak concepts are strengthened in parallel, instead of the whole class stopping or the weak student being left behind.
| Topic-based | Concept-based | |
|---|---|---|
| Unit of progress | Chapter completed | Concept secured |
| What a failure tells you | Which chapter to revise | Which idea to fix, and which earlier idea it rests on |
| Evidence used | A test score at the end | Every response, grouped by the concepts behind it |
| Typical remedy | Re-teach the chapter | Strengthen the earliest concept that isn't secure |
| Risk it avoids | — | Treating a symptom two years downstream of the cause |
Why it is rarely done properly
Concept-based learning is not a new idea, and most teachers already think this way instinctively. The obstacle has always been that doing it at scale requires three things at once, and the third is brutal.
A concept map for the subject
Every concept identified, and every dependency between them written down — for each subject and each class.
Questions tagged to concepts
Not tagged to the chapter they appear in, but to the ideas they actually require. One question usually needs several.
Continuous analysis
Thirty students × hundreds of responses a week × dozens of interlinked concepts. This is the part no person can do by hand.
The first two are careful editorial work. The third is an information problem — and it is precisely the part that software should be taking off a teacher's hands. It is also why a concept-based approach and an AI-powered one tend to arrive together: the map makes the reasoning possible, and the engine makes it affordable.
How Sublogy does it
Sublogy is built concept-first from the bottom up.
- Questions are linked to concepts, not chapters. Every question carries the set of ideas it actually depends on, so a response is evidence about several concepts at once.
- Concepts are linked to their prerequisites. That map is what allows a weak concept to be traced downwards to the earliest one that isn't secure.
- No conclusion is drawn from a single answer. A single incorrect answer is weak evidence — a slip, a misread, a distraction. A concept is assessed on the pattern across many questions that share it.
- Practice targets the foundation, not the symptom. If fractions are the real problem, the student gets fractions — even though the question they failed was about discounts.
- Every question carries a worked example. Not just the answer, but the steps, written the way a teacher would explain them.
- Progress is reported per concept. A teacher sees which concepts need attention and how many students each one affects, rather than a column of percentages.
The engine doing that work is CUBALE, the Concept Understanding Based Adaptive Learning Engine. The concept map is what it reasons over; adaptive learning is what it produces.
One honest caveat. A concept map is a model, and models are approximations. CUBALE identifies patterns and detects likely gaps; it does not claim to know every misconception perfectly. Recommendations are shown to a teacher as something worth looking at, not as a verdict — because the person in the room still knows things the data cannot see.
Questions about concept-based learning
What is concept-based learning?
Concept-based learning organises teaching, practice and measurement around individual concepts — single transferable ideas such as 'a percentage is a fraction out of a hundred' — rather than around chapters, topics or test scores. Because concepts depend on other concepts, it also makes it possible to trace a difficulty back to the earlier idea causing it.
How is a concept different from a topic?
A topic is a unit of the syllabus, such as 'Profit, Loss and Discount'. A concept is one idea inside it, such as 'a discount is a share of the marked price'. A topic can contain a dozen concepts, and a student can be secure on most of them and stuck on one — which a topic-level score cannot show you.
Why does concept-based learning matter for weaker students?
Because a topic-level score tells a struggling student only that they are behind, while a concept-level view tells them exactly which idea to fix. It also prevents the most common wasted intervention: re-teaching the topic where the mistake appeared instead of the earlier concept that actually caused it.
Does concept-based learning replace the syllabus?
No. The syllabus still sets what has to be covered and when. Concept-based learning changes the resolution at which progress is tracked underneath it, so a student can be moved forward through the syllabus while a specific weak concept is strengthened in parallel.
How does Sublogy use concept-based learning?
Every question in Sublogy is linked to the concepts it depends on, and those concepts are linked to the earlier concepts they are built on. CUBALE, Sublogy's adaptive engine, assesses each concept on the weight of evidence across many questions and uses that map to find the earliest idea that isn't secure.
Concept-level progress, for every student in the class
Sublogy is provided through schools, for classes 6 to 10.
Sublogy for schools How CUBALE works