Explanation changes learning because it forces relationships to become explicit. Searches for self-explanation, explanation in teaching, how to explain concepts, worked examples, explanatory learning and why explaining helps you learn all point to a practical distinction: recognising an answer is not the same as being able to say why it is true, how it works, what evidence supports it or when it stops applying.
Explanation is not automatically deep. Learners can repeat a teacher’s words without reconstructing the relationship, and teachers can provide so much explanation that students have little left to think through. Effective explanation therefore balances clarity with learner generation, examples with principles, and support with a later requirement for independent reconstruction.
This guide explains self-explanation, teacher explanation, worked examples, causal reasoning, analogies, diagrams, retrieval and transfer. Its central proposition is simple: explanation changes learning when a learner can reconstruct the relationship behind an answer and use that relationship again.
Your 50-second route
Ask what relationship needs explaining. Supply missing knowledge. Model one clear example. Ask the learner to explain why each important step follows. Contrast with a non-example. Remove the model. Require reconstruction. Change the surface form and test whether the principle still controls the answer. If the explanation collapses into memorised wording, return to the underlying relationship.
Expandable contents — explanation, examples, subjects and transfer
1. Explanation makes relationships explicit · 2. Knowing versus explaining · 3. Self-explanation · 4. Teacher explanation · 5. Worked examples · 6. Why questions · 7. How questions · 8. Causal explanation · 9. Mechanistic explanation · 10. Comparative explanation · 11. Examples and non-examples · 12. Analogies · 13. Diagrams · 14. Vocabulary · 15. Prior knowledge · 16. Working memory · 17. Cognitive load · 18. Retrieval · 19. Feedback · 20. Mistakes · 21. Peer explanation · 22. Discussion · 23. Reading · 24. Writing · 25. Mathematics · 26. Science · 27. Primary learners · 28. Secondary learners · 29. Parents · 30. Teachers · 31. Overexplaining · 32. Illusion of explanatory depth · 33. AI explanations · 34. Seven-day explanation experiment · 35. Thirty-day review · 36. Explanation and transfer · 37. Explanation and critical thinking · 38. Explanation and memory · 39. Explanation and agency · 40. World-return test
Useful routes: Why Discussion Changes Learning, Why Writing Changes Thinking, Why Questions Change Thinking, Why Feedback Changes Learning and the How X Works library.
1. Explanation makes relationships explicit
Explanation makes relationships explicit by requiring the learner to connect events, concepts, steps or evidence rather than merely name them. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
2. Knowing versus explaining
A learner can recognise an answer without being able to explain why it is correct, so explanation can reveal hidden gaps in understanding. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
3. Self-explanation
Self-explanation is useful when learners connect steps to principles rather than simply narrating what is visible on the page. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
4. Teacher explanation
Teacher explanation should reduce unnecessary search while preserving enough learner work for understanding and later independence. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
5. Worked examples
Examples become more informative when contrasted with non-examples that expose the boundary of the concept. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
6. Why questions
Analogies help when the mapped relationship is clear and their limits are stated before surface similarities become misconceptions. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
7. How questions
Explanation changes learning most usefully when the learner can later reconstruct the relationship and apply it to a changed case. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
8. Causal explanation
Explanation makes relationships explicit by requiring the learner to connect events, concepts, steps or evidence rather than merely name them. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
9. Mechanistic explanation
A learner can recognise an answer without being able to explain why it is correct, so explanation can reveal hidden gaps in understanding. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
10. Comparative explanation
Self-explanation is useful when learners connect steps to principles rather than simply narrating what is visible on the page. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
11. Examples and non-examples
Teacher explanation should reduce unnecessary search while preserving enough learner work for understanding and later independence. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
12. Analogies
Examples become more informative when contrasted with non-examples that expose the boundary of the concept. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
13. Diagrams
Analogies help when the mapped relationship is clear and their limits are stated before surface similarities become misconceptions. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
14. Vocabulary
Explanation changes learning most usefully when the learner can later reconstruct the relationship and apply it to a changed case. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
15. Prior knowledge
Explanation makes relationships explicit by requiring the learner to connect events, concepts, steps or evidence rather than merely name them. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
16. Working memory
A learner can recognise an answer without being able to explain why it is correct, so explanation can reveal hidden gaps in understanding. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
17. Cognitive load
Self-explanation is useful when learners connect steps to principles rather than simply narrating what is visible on the page. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
18. Retrieval
Teacher explanation should reduce unnecessary search while preserving enough learner work for understanding and later independence. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
19. Feedback
Examples become more informative when contrasted with non-examples that expose the boundary of the concept. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
20. Mistakes
Analogies help when the mapped relationship is clear and their limits are stated before surface similarities become misconceptions. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
21. Peer explanation
Explanation changes learning most usefully when the learner can later reconstruct the relationship and apply it to a changed case. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
22. Discussion
Explanation makes relationships explicit by requiring the learner to connect events, concepts, steps or evidence rather than merely name them. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
23. Reading
A learner can recognise an answer without being able to explain why it is correct, so explanation can reveal hidden gaps in understanding. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
24. Writing
Self-explanation is useful when learners connect steps to principles rather than simply narrating what is visible on the page. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
25. Mathematics
Teacher explanation should reduce unnecessary search while preserving enough learner work for understanding and later independence. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
26. Science
Examples become more informative when contrasted with non-examples that expose the boundary of the concept. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
27. Primary learners
Analogies help when the mapped relationship is clear and their limits are stated before surface similarities become misconceptions. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
28. Secondary learners
Explanation changes learning most usefully when the learner can later reconstruct the relationship and apply it to a changed case. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
29. Parents
Explanation makes relationships explicit by requiring the learner to connect events, concepts, steps or evidence rather than merely name them. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
30. Teachers
A learner can recognise an answer without being able to explain why it is correct, so explanation can reveal hidden gaps in understanding. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
31. Overexplaining
Self-explanation is useful when learners connect steps to principles rather than simply narrating what is visible on the page. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
32. Illusion of explanatory depth
Teacher explanation should reduce unnecessary search while preserving enough learner work for understanding and later independence. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
33. AI explanations
Examples become more informative when contrasted with non-examples that expose the boundary of the concept. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
34. Seven-day explanation experiment
Analogies help when the mapped relationship is clear and their limits are stated before surface similarities become misconceptions. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
35. Thirty-day review
Explanation changes learning most usefully when the learner can later reconstruct the relationship and apply it to a changed case. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
36. Explanation and transfer
Explanation makes relationships explicit by requiring the learner to connect events, concepts, steps or evidence rather than merely name them. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
37. Explanation and critical thinking
A learner can recognise an answer without being able to explain why it is correct, so explanation can reveal hidden gaps in understanding. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
38. Explanation and memory
Self-explanation is useful when learners connect steps to principles rather than simply narrating what is visible on the page. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
39. Explanation and agency
Teacher explanation should reduce unnecessary search while preserving enough learner work for understanding and later independence. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
40. World-return test
Examples become more informative when contrasted with non-examples that expose the boundary of the concept. The important evidence is not the length of the explanation but whether it captures the governing relationship accurately. A concise explanation can be strong when every necessary link is present; a long explanation can remain weak when it circles the topic without identifying why one step leads to the next.
Consider a fictional learner who correctly solves a problem but says, “I just knew which formula to use.” Ask which feature of the problem made that formula relevant. If the learner cannot answer, the final result may conceal a method-selection gap. Explanation makes the hidden decision visible. Teaching can then connect the cue, the principle and the method rather than simply celebrating the correct number.
Begin with enough knowledge for explanation to be possible. Asking a novice to discover an unfamiliar mechanism through repeated “why?” questions can create guessing. Supply the essential concept or worked example, then ask the learner to connect the parts. As knowledge grows, reduce the amount supplied and increase the amount the learner must reconstruct independently.
Use contrasts to sharpen the explanation. If a learner says that evaporation means water disappearing, compare a situation that fits the scientific concept with one that does not. If a learner says every percentage increase can be cancelled by the same percentage decrease, use transparent numbers to expose the changing base. A good non-example reveals which condition the explanation must include.
After teaching, remove the model and ask for the relationship in the learner’s own words. Accept accurate paraphrase rather than demanding the teacher’s exact sentence. Then use a changed example. If the learner reproduces the wording but cannot recognise the principle elsewhere, the explanation has not yet become flexible knowledge. Return to the comparison between examples and ask what remains invariant.
For parents, “Why does that answer make sense?” can be useful when the child has enough knowledge to respond; it should not become an interrogation after every item. For teachers, explanation should target the bottleneck rather than repeat the entire lesson. For students, explaining aloud or in writing can expose the precise sentence where understanding becomes vague. That sentence is often a better study target than the whole chapter.
AI-generated explanations create a new verification problem. Fluency can make an explanation sound complete even when a condition is missing or a claim is inaccurate. Ask the learner to check important claims against reliable sources, identify the principle in their own words and solve a fresh example without relying on the generated text. The tool can support explanation; it should not erase the evidence that the learner can explain.
The world-return test is whether explanation improves later decisions. Can the learner select the method, justify the inference, identify the scientific mechanism or revise the argument when the original example disappears? If yes, explanation has become part of capability. If not, locate the missing relationship and rebuild it with a clearer example, contrast or prerequisite.
Research floor and further routes
Research starting points include Institute of Education Sciences evidence resources and APA principles for learning and teaching. Effects of explanation depend on prior knowledge, prompts, examples, feedback and the criterion used to measure learning. Continue through writing, discussion, knowledge and practice.