MindOS · Explanation State · State → Why → Because → Evidence → Check → Reconstruct → Transfer
Wait, What? A Student Can Produce the Correct Step and Still Not Know Why It Belongs There
A learner writes the right formula. The answer is correct. Everyone moves on.
Then the next question changes slightly and the method disappears.
The first success may have been procedural recognition rather than owned reasoning. One way to expose that difference is to ask for an explanation: Why does this step make sense here?
Quick Answer
Explanation State asks whether the learner can make the hidden relation explicit: why a step follows, why evidence supports a claim, why a variable matters, why a method applies, or why an answer is wrong.
DO / CLAIM / ANSWER ↓ WHY? ↓ NAME THE RELATION ↓ CONNECT TO EVIDENCE OR PRINCIPLE ↓ CHECK FOR CONTRADICTION ↓ RECONSTRUCT WITHOUT THE MODEL ↓ EXPLAIN IN A NEW CASE
The Owned Job of This Page
This page owns self-explanation as a learner operation. Comparison asks what is similar or different. Elaboration asks how new knowledge connects outward to prior knowledge and examples. Explanation asks the narrower question: can the learner make the internal logic of this step or relationship explicit?
What Self-Explanation Can Reveal
- The learner knows a rule but not its condition.
- They copied a procedure without understanding the causal relation.
- They can state evidence but cannot connect it to the claim.
- They use the correct scientific keyword without reconstructing the mechanism.
- They can repeat the teacher’s sentence but cannot generate their own explanation.
- They can explain one familiar example but not a changed one.
How Do We Know Explanation Can Help?
Self-explanation has a substantial research history. A 2025 three-level meta-analysis in Educational Psychology Review synthesised 204 effect sizes from 56 studies in digital learning environments and found a positive overall effect on academic performance, including retention and transfer, while also showing that effects vary with learner, material and learning-environment characteristics. See Enhancing Academic Performance Through Self-Explanation in Digital Learning Environments.
The useful teaching conclusion is not “make students explain everything.” It is that prompts to explain can turn passive inspection into constructive processing and can expose gaps that polished answers hide.
Discrimination Test 1: Explanation or Paraphrase?
A learner may simply restate the sentence with different words. That is not always explanation.
Ask for the link:
- Why does this evidence support the claim?
- Why does this operation follow from the relationship?
- Why does changing this variable alter the outcome?
- Why would the answer be different if this condition changed?
If the learner can only rename the statement, the causal or logical connection may still be missing.
Discrimination Test 2: Explanation or Memorised Script?
Change the example but preserve the relation. Ask the learner to explain again without the original wording.
If the explanation survives, the reasoning may be becoming generative. If the exact teacher phrase is required, the state is still fragile.
Discrimination Test 3: Missing Explanation or Missing Knowledge?
A learner cannot explain photosynthesis if the foundational concepts are missing. Explanation prompts do not replace instruction.
Ask whether the necessary terms, facts and relations are retrievable first. If not, route back to concept instruction or Retrieval State.
Intervention 1: Ask “Why This?” Instead of “Do You Understand?”
“Do you understand?” often invites a yes/no answer. Replace it with an observable operation.
- Why is this method appropriate?
- Why does this evidence matter?
- Why is this sentence an inference rather than a direct detail?
- Why did the previous route fail?
- Why would this rule stop working here?
Intervention 2: Explain the Step Before Seeing the Next Step
Worked examples are most useful when learners process the route rather than watch it. Pause after a step and ask what the step achieved, why it was legal, and what should happen next.
This keeps explanation connected to the actual reasoning sequence and links naturally to Worked Example State.
Intervention 3: Make the Learner Explain an Error
“This is wrong” identifies an outcome. “This fails because…” identifies a mechanism.
Ask the learner to locate the first invalid step, explain why it fails, state the principle that was violated, and reconstruct the route. This connects explanation to Feedback State.
Intervention 4: Fade the Explanation Prompt
TEACHER: Why did you do that? ↓ PROMPT CARD: Why? Evidence? Condition? ↓ LEARNER PAUSES BEFORE STEP ↓ LEARNER SELF-PROMPTS ↓ LEARNER EXPLAINS ONLY WHEN NEEDED ↓ LEARNER USES EXPLANATION TO CHECK A NEW CASE
The aim is not endless verbalisation. Mature learners should be able to invoke explanation when uncertainty, novelty or error makes it useful.
Explanation in English, Mathematics and Science
- English: explain why quoted evidence supports an inference or why a writer’s word choice changes tone.
- Mathematics: explain why an operation preserves equivalence, why a method applies, or why an answer is unreasonable.
- Science: explain mechanism, variable relationships, causal chains and why evidence supports one interpretation over another.
Evidence Boundary
Self-explanation is not automatically beneficial in every form. Weak explanations can rehearse misconceptions. Excessive prompting can overload a novice. Some material is better served by direct teaching first. Feedback and task design matter.
MindOS therefore uses explanation as a diagnostic and constructive operation, not as a universal command to “talk more.”
Transfer Test
- Change the numbers but preserve the relation.
- Change the representation.
- Ask for the same explanation in an unseen passage or experiment.
- Present a plausible wrong explanation and ask for repair.
- Ask what would have to change for the explanation to become false.
If the learner can explain the underlying relation after surface changes, Explanation State is moving toward transfer-ready.
Examination Implication
Many examination questions reward compressed explanation: cause → mechanism → consequence; claim → evidence → inference; method → condition → result. Practising explicit explanation can help build those structures, but examination readiness also requires producing them accurately and efficiently under time.
Return Test: What Came Back?
- Can the learner explain without the tutor’s exact wording?
- Can they identify the relation rather than merely paraphrase?
- Can they explain an error?
- Can they decide when an explanation is needed?
- Does explanation survive a new example?
Parent and Tutor Teaching Guide
When a child gets an answer right, occasionally ask one more question: Why does that work? When they get it wrong, ask: Where does the explanation first stop making sense?
Use the answer diagnostically. If the learner cannot explain because knowledge is absent, teach. If the knowledge is present but the relation is hidden, help them make the relation visible.
MindOS Direction Graph
EXPLANATION STATE ├── Cannot retrieve facts? → RETRIEVAL ├── Cannot see relation? → REPRESENTATION / COMPARISON ├── Paraphrases only? → ASK WHY / BECAUSE ├── Repeats script? → CHANGE EXAMPLE ├── Explanation incorrect? → FEEDBACK / CONCEPT REPAIR ├── Prompt dependence? → FADE └── Explains new case independently? → TRANSFER
Continue Through MindOS
MindOS boundary: Self-explanation is an educational learning operation. Difficulty explaining a concept is evidence about the present task state, not a clinical or psychological diagnosis.
