Mathematics Dependency Evidence | Validating Prerequisite Links
Mathematics ideas are connected, but a connection should not be treated as absolute merely because it sounds plausible. This reader page explains how we think about the strength of learning relationships.
What evidence can support a connection?
- The mathematical structure itself.
- Singapore curriculum progression and official syllabus expectations.
- Relevant education research.
- Repeated patterns in student work.
- Cases that contradict the proposed relationship.
Some earlier ideas are genuine prerequisites. Others simply make later work easier. Good teaching keeps that distinction open to evidence.
A learning map is useful only if it is willing to redraw a road when reality disagrees.
PHASE 4 · DEPENDENCY EVIDENCE READER GUIDE
Quick Read: when is one Mathematics idea genuinely a prerequisite for another?
A dependency deserves strong confidence when the earlier capability is structurally required, its absence reliably predicts failure in the later task, and repairing it improves later performance. Many useful learning connections are weaker than that and should be described accordingly.
Mathematics is richly connected, but not every connection is a hard prerequisite. Some ideas are necessary. Some are strong facilitators. Some usually appear earlier because the curriculum is organised that way. Others simply tend to develop together. Treating every arrow as equally strong creates unnecessary remediation and can send learners backward when they already have enough capability to proceed.
One-sentence answer: a dependency map should distinguish what is required from what is merely helpful, typical or correlated—and remain willing to redraw the edge when evidence disagrees.
Not every edge means the same thing
| Relationship | Meaning | Example question |
|---|---|---|
| Required prerequisite | The later task cannot be performed reliably without the earlier capability. | Can equation transformation work if equality itself is not understood? |
| Strong facilitator | The earlier capability lowers load or improves transfer but is not always strictly necessary. | Does fast mental arithmetic help algebra even when written arithmetic can compensate? |
| Representation bridge | The earlier form helps the learner access or translate the later form. | Does a ratio table support transition into y = kx? |
| Curriculum precursor | The topic is taught earlier by design, but the order alone does not prove necessity. | Is this sequence pedagogically convenient or logically required? |
| Common co-development | Capabilities often improve together without one necessarily causing the other. | Are stronger graph skills and stronger algebra both consequences of broader function understanding? |
These distinctions protect the graph from becoming a one-way remedial staircase where every difficulty is blamed on the earliest possible topic.
Five evidence sources should converge
- Mathematical structure. Does the later idea logically use the earlier relationship?
- Curriculum progression. Is the earlier capability expected before the later one, and for what educational reason?
- Research. Is there credible evidence that the earlier capability supports later learning or transfer?
- Student work. Does weakness in the earlier capability actually appear when the later task fails?
- Counterexamples. Are there learners who perform the later task reliably despite weakness in the supposed prerequisite?
No single source automatically settles every edge. Mathematical structure can establish a logical relation, but not necessarily the best teaching sequence. Curriculum order can show expected progression, but not prove causality. Student work can be highly relevant to an individual learner, but one case may be noisy.
Three examples of edge strength
Equality → equation solving
This is close to a structural prerequisite. Solving equations requires transformations that preserve equality. A learner may memorise “move across, change sign,” but unreliable equality reasoning makes the method fragile when the equation changes form.
Mental arithmetic speed → algebra
Fast arithmetic can reduce working-memory load, but it is not necessarily a hard prerequisite. A learner with slower mental arithmetic may still perform algebra strongly using accurate written methods. The edge may therefore be facilitative rather than absolute.
Bar model → algebraic problem solving
A bar model can be a powerful representation bridge, but algebra does not require permanent bar-model use. Once the learner can reconstruct the relationship symbolically, the representation should become optional rather than mandatory.
Counterexamples are not nuisances—they refine the edge
Suppose the map predicts that capability A is required for capability B. A learner then performs B independently despite weakness in A. Several possibilities follow:
- The edge was too strong and should be downgraded from prerequisite to facilitator.
- The learner is compensating through another route.
- The probe for A measured the wrong thing.
- The later task B did not actually demand the suspected dependency in that form.
The correct response is to investigate—not to discard the learner because the learner failed to behave like the map.
A good graph treats counterexamples as instructions for improving the graph.
Confidence should be updated, not frozen
| New evidence | Reasonable update |
|---|---|
| Repeated learner failures in B coincide with weakness in A, and repairing A improves B. | Increase confidence in the dependency. |
| Learners often perform B despite weakness in A. | Downgrade the edge or narrow its conditions. |
| A helps only under high load or timed work. | Record the edge as conditional rather than universal. |
| Research and classroom evidence conflict. | Preserve uncertainty and inspect context before making a stronger claim. |
| A new representation bypasses A successfully. | Consider multiple routes to B rather than one mandatory path. |
This is more useful than binary “true/false” thinking because educational dependencies often depend on task form, stage, representation and available compensating skills.
What parents should understand about “foundational gaps”
Foundational repair can be extremely valuable, but the phrase should not justify indefinite backward teaching. A useful foundation is one whose improvement changes the learner’s current Mathematics.
- Ask which current task the earlier skill is supposed to support.
- Ask what evidence shows that the earlier skill is actually limiting performance.
- Ask how the repair will reconnect to current work.
- Ask what would count as enough repair to move forward.
- Ask whether another route could support the learner more efficiently.
The purpose of a dependency graph is not to justify more remediation. It is to identify the smallest earlier repair that restores forward motion.
Frequently asked questions
Does “taught earlier” mean “prerequisite”?
No. Curriculum sequence is evidence about expected progression, but necessity should also be supported by mathematical structure and learner performance.
Can there be more than one route into a later topic?
Yes. Learners can sometimes compensate through stronger representations or alternative strategies. A good map should allow multiple routes where the Mathematics permits them.
When should an edge be removed?
When repeated evidence shows that the proposed relationship is not useful or is misleading. Sometimes an edge should instead be weakened, narrowed to particular conditions or reclassified.
What is the strongest evidence for a practical dependency?
When weakness in the earlier capability predicts a specific later failure, a targeted repair improves that later performance, and the improvement transfers beyond the original practice surface.
The larger idea: a learning graph is a hypothesis about how capability carries load
The value of a dependency graph is not the appearance of completeness. Its value is that it makes assumptions visible: this earlier capability is believed to support that later one, under these conditions, with this degree of confidence.
Once the claim is visible, it can be tested. That is what turns a diagram of topics into a useful educational instrument.
A trustworthy learning map is not the map with the most arrows. It is the map whose arrows remain answerable to Mathematics and to learners.