Self-Explanation: A Study Technique for Learning Concepts, Not Just Terms

You can recognize every word in a textbook paragraph and still be unable to explain what the paragraph means. Self-explanation is a way to catch that gap while you are studying. Instead of moving through a page silently, you pause and explain what each idea means, how it connects to what came before, and what follows from it.

The goal is not to recite the sentence. It is to build a usable model. If a text says that blood moves through a chamber before entering a vessel, a self-explanation might ask what pressure changes, why the direction matters, and what would happen if the connection were blocked.

This technique has a study behind it, along with a limitation: the sample was small and the lesson was specific. It is worth understanding, not turning into a universal promise.

What the original study found

Chi, de Leeuw, Chiu, and LaVancher studied self-explanation while eighth-grade students learned about the human circulatory system. In their 1994 Cognitive Science paper, 14 students were prompted to explain each line of a passage aloud. A control group of 10 students read the passage twice without the same explanation prompts.

The students completed pretests and posttests. The researchers looked not only at recall of stated information but also at questions requiring inferences about how the system worked. The prompted group showed greater improvement. Within that group, students who generated many explanations showed deeper understanding than students who generated fewer. The authors reported that all of the high-explaining students formed an accurate mental model of the circulatory system, while many control-group and low-explaining students did not.

The design does not show that saying more words automatically causes learning. The self-explanation group was tiny, the topic was one science passage, and the comparison was reading twice without prompts. Students were not randomly sampled from every age and subject. The result is best used as a mechanism-informed study idea: explanations may help when they make you connect new information with prior knowledge and generate inferences.

What self-explanation sounds like

A self-explanation can be simple. After each sentence or short chunk, ask one question and answer it in your own words:

• What does this mean? Translate technical language into plain language.
• Why is this true? Connect the claim to a cause, rule, or observation.
• How does it fit? Link it to the previous step or a larger structure.
• What would follow? Make a prediction or inference.
• What would change it? Name a condition, exception, or boundary.

Suppose a chemistry text says that increasing temperature can increase the rate of a reaction. A weak explanation repeats the sentence: “Higher temperature increases the rate.” A stronger one might say: “The particles have more kinetic energy, so a larger share of collisions can reach the activation energy. This does not mean every reaction speeds up in the same way; the effect depends on the reaction and conditions.”

A four-step self-explanation routine

1. Read a small chunk

Use one sentence, a short paragraph, a diagram label, or one step in a worked example. Large chunks encourage vague summaries because you cannot tell which idea you failed to understand.

2. Hide the text briefly

Look away or cover the next line. Explain the chunk aloud or write two sentences. If you need to quote the source, stop and translate it. The point is to produce an explanation, not to keep your eyes on the answer while paraphrasing.

3. Connect and predict

State how the chunk relates to what came before. Then make one prediction, example, or counterexample. For a process, describe the next step. For a definition, give an instance and a non-instance. For an argument, state what evidence would support or challenge it.

4. Check the source

Reopen the text. Mark three different outcomes:

• Accurate and connected: keep moving.
• Incomplete: add the missing relationship or condition.
• Wrong: write a corrected explanation and identify the assumption that led you astray.

An explanation that sounds smooth but contradicts the source is not a success. Checking prevents fluency from becoming false confidence.

Worked examples across subjects

Anatomy or physiology

Text: “The pulmonary artery carries blood from the right ventricle to the lungs.”

Self-explanation: “An artery is defined by carrying blood away from the heart, not by oxygen content. This vessel carries relatively deoxygenated blood, so ‘artery’ does not always mean oxygen-rich. The right ventricle sends it toward the lungs, where gas exchange occurs.”

Follow-up: “What would happen to oxygen delivery if this route were blocked?”

History

Text: “A policy increased tensions between the two groups.”

Self-explanation: “The policy is not just a date to memorize. I need to identify which rule changed, who bore the cost, and why the affected group interpreted it as a threat. The tension should be connected to a mechanism, such as lost autonomy or unequal resources.”

Follow-up: “Which evidence would show that the policy, rather than another event, contributed to the tension?”

Mathematics

Worked example: A solution changes both sides of an equation by the same amount.

Self-explanation: “The operation preserves equality because applying the same inverse operation to both sides keeps the two expressions balanced. If I apply it to only one side, the new equation no longer represents the original relationship.”

Follow-up: “Can I create a number example where doing this reveals the unknown?”

Prompts that produce reasoning

Generic prompts produce generic summaries. Keep a small set of prompts nearby and choose one that matches the material:

• “What problem is this step solving?”
• “What must be true for this rule to apply?”
• “What is the causal link between these two facts?”
• “How would I recognize this in a new example?”
• “What is easy to confuse with this, and what separates them?”
• “What information is implied but not stated?”

Do not use every prompt on every line. One good question is enough to expose a missing link. Rotate prompts when the material changes: mechanisms need causal questions, categories need boundary questions, and procedures need next-step questions.

How to avoid turning it into a rewriting task

Self-explanation can fail in predictable ways.

Paraphrase without a connection. Changing the words while preserving no relationship creates a shorter copy, not an explanation. Add “because,” “therefore,” or “this matters when” and complete the thought.

Explain everything equally. Some lines define a term; others carry the mechanism. Spend more time on transitions, exceptions, and steps that change the outcome.

Keep the source visible. If your eyes never leave the page, you may be recognizing the wording. Hide a small chunk, produce the explanation, then check it.

Skip correction. A confident but wrong explanation is more dangerous than an obvious blank. Compare with the source and record the repair.

Overload the session. A paragraph-by-paragraph explanation can be tiring. Start with ten minutes on one difficult section. Continue only if the explanations are still specific.

Pair self-explanation with retrieval and cards

Self-explanation and retrieval practice overlap but are not identical. Retrieval asks you to produce knowledge with the source closed. Self-explanation asks you to make relationships explicit while learning and checking them. Use both in sequence:

1. Read a bounded section.
2. Explain the key steps and connections aloud.
3. Close the source and reconstruct the whole structure.
4. Check the reconstruction and repair gaps.
5. Return later with a few focused prompts.

When turning study material into flashcards, preserve the reasoning. A basic card can ask for a mechanism; a cloze can hide the condition that makes a rule apply; a scenario card can ask you to predict the next step. The Cloze vs. Basic guide discusses how context changes the usefulness of a card. The broader From PDF Highlights to Quality Flashcards workflow is useful when your source contains dense explanations rather than isolated facts.

A card should not ask you to reproduce every sentence you explained. Keep the durable relationship and enough context to retrieve it. For example:

• Weak: “What is the pulmonary artery?”
• Stronger: “Why is the pulmonary artery still called an artery even though it carries deoxygenated blood?”
• Application: “If blood cannot leave the right ventricle through the pulmonary artery, which downstream function is disrupted first, and why?”

The stronger prompts test category boundaries and mechanisms. They do not replace diagrams, worked problems, or full explanations, but they make the conceptual links easier to revisit.

A 12-minute practice plan

Use this for one section tonight:

1. Two minutes: Preview the heading, diagram, and key terms.
2. Four minutes: Read one small chunk at a time and explain each important transition.
3. Two minutes: Close the source and draw or state the structure from memory.
4. Two minutes: Check for missing or wrong links.
5. Two minutes: Write two retrieval prompts for tomorrow.

Measure the quality of the explanations, not their length. A short explanation that identifies the right cause and condition is better than a fluent paragraph that merely repeats vocabulary.

What this study does and does not establish

Chi and colleagues' study is a strong reason to try self-explanation, not a reason to claim that every learner, subject, or prompt will respond the same way. Fourteen prompted eighth-graders and ten controls are not enough to estimate a universal effect. The task focused on a circulatory-system passage, and the researchers compared prompted explanation with rereading rather than with every active-learning alternative.

The study also found differences within the prompted group: students who explained more showed deeper understanding. That observation is consistent with a useful caution. Self-explanation is not a magic phrase. The explanation must connect ideas, generate inferences, and be checked against the material.

Make the explanation earn its time

Self-explanation turns “I recognize this paragraph” into a more demanding question: “Can I say what this means, why it follows, and what would happen next?” The original study found better conceptual understanding for a small group of eighth-graders prompted to explain a circulatory-system text, especially among students who generated many explanations.

Try the technique on small chunks, hide the source before explaining, ask for a connection or prediction, and correct the result. Keep the sample and subject limits in view. The value is not in saying more; it is in making the relationships in the material explicit enough to retrieve, test, and use.