把一条学习原理，连到你能改的那一处设计。

Self-explanation and deep explanatory questioning

Ask learners to explain why a fact, step, representation, or causal link makes sense and how it connects to prior knowledge.

Recommendation

Retrieve when explanation quality can be observed and corrected.

Mechanism

Generating explanations exposes gaps, integrates new information with schemas, and makes relations available for inference.

Applicable contexts and prerequisites

Learner Stages

• primary
• secondary
• adult

Task Types

• concept integration
• worked-example study
• causal reasoning

Contexts

• individual reflection
• peer explanation
• guided questioning

Lifespan Stages

• primary
• middle
• secondary
• postsecondary

• Sufficient domain knowledge to generate a meaningful explanation
• A prompt aimed at a consequential relation

Constraints and risks

• Provide feedback or models when explanations can entrench misconceptions
• Avoid generic why prompts for every step

• Fluent but incorrect explanations
• Prompt overload
• Scoring verbosity rather than conceptual relations

Observable evidence

• Indicator

  Integrated and revisable explanation

  Measure

  Presence and correctness of causal, structural, or principle-based links plus transfer

  Timing

  During learning and on a novel application

  Caveat

  Longer explanations are not necessarily better

Evidence

context-dependent

Evidence-rated guidance supports deep explanatory questions; effectiveness depends on knowledge, prompt, and feedback.

Confidence

moderate

Directness

mixed-direct-and-inferential

Generalizability

Broad mechanism with domain-specific implementation.