Learning concept map: from cognitive supports to mastery
knowledge record: Learning concept map: from cognitive supports to mastery.
Chinese version: /zh-hant/knowledge/learning-concept-map/
Status: Working map of the concepts discussed so far, reframed on 2026-09-20. Use it to organise diagnosis and practice; it is not a final theory of learning.
Created: 2026-09-11
One connected model
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Cognitive supports ── enable / constrain ──┐
│
World model / knowledge ───────────────────┼─> Understanding ability
│ ├─ explain and predict
New information + feedback ── expose ───────┤ ├─ locate failed assumptions
model mismatch ── revise + integrate ───────┤ ├─ apply / transfer
│ └─ test the reconstructed model
Purposeful practice ── strengthens ─────────┘
│
└─> reliable performance under defined conditions = Mastery
├─ recognition
├─ explanation
├─ guided application
├─ independent proficiency
├─ transfer / flexible mastery
└─ creation / teaching
The concepts and their links
1. Cognitive supports
Attention, working memory, and executive control / self-regulation help a learner focus, hold conditions in mind, plan, monitor, and change strategy. They affect how easily a learner can process and express ideas.
They are supports, not proof of understanding. Strong memory or concentration can coexist with definition-only recall.
2. Knowledge as a world model
Knowledge is the relatively dependable, usable part of the learner’s internal world model: organised representations of facts, vocabulary, examples, relations, mechanisms, conditions, and procedures. External books, notes, data, and LLM replies are model resources, not automatically the learner’s knowledge.
Learning does not merely add resources. It tests the current model, identifies the smallest failed relation or assumption, then revises and integrates that part with what remains accurate. A weak or incomplete model leads to omitted conditions, incorrect causal inferences, and weak transfer.
3. Understanding ability
Understanding is demonstrated by using a model rather than simply repeating information. The current working tests are:
- Explain a relationship or mechanism.
- Predict what changes when conditions change.
- Apply / transfer the relation to a new case.
- Diagnose, revise, and integrate an assumption after error or counterevidence.
This makes understanding a bridge between stored knowledge and performance.
4. Mastery
Mastery is reliable, independent, and appropriately flexible performance for a defined class of tasks. It normally draws on understanding, but it also needs practice, procedural fluency, feedback, and correction.
A learner may understand a principle but lack mastery if performance is slow, inconsistent, prompt-dependent, or limited to one familiar example.
5. Mastery levels
| Level | Core evidence |
|---|---|
| Recognition | Identifies a word, rule, method, or example. |
| Understanding / explanation | Explains the main relation or reason. |
| Guided application | Uses the idea with a model, prompt, or worked example. |
| Independent proficiency | Performs typical tasks accurately and repeatedly without prompts. |
| Transfer / flexible mastery | Adapts the idea to unfamiliar, changed, or complex situations. |
| Creation / teaching | Diagnoses errors, compares methods, improves an approach, and teaches it. |
6. Evidence criteria
Do not infer a level from confidence or one correct answer. Check:
- Accuracy — is the result correct?
- Stability — is it repeatable?
- Independence — does it require prompts or notes?
- Transfer — does it survive a changed situation?
- Reflection — can the learner find and repair a relevant error?
7. Model revision and reconstruction loop
Current model → changed case, new information, or feedback → identify mismatch → locate earliest failed assumption → revise and integrate the relation → test in a new case
Revision is not wholesale rejection of prior knowledge. Preserve relations that remain accurate; narrow, replace, or reorganise only the relation, variable, mechanism, or boundary condition that failed.
8. Deliberate-practice loop
Goal → diagnose → narrow drill → attempt → feedback → reflection → next drill → review
The loop improves both the learner’s model and the reliability of their performance. Feedback reveals the earliest failed step; targeted practice then strengthens a specific weak link and tests whether the revision transfers.
Diagnostic questions
When a learning task goes badly, separate possible causes:
- Is necessary knowledge, vocabulary, or a conceptual model missing?
- Is attention, working memory, or self-regulation overloaded?
- Is the learner at a lower mastery level than the task demands?
- Is a condition or causal variable being treated as fixed when it is not?
- What smallest drill would distinguish these explanations?