Self-Explanation in Learning

Self-explanation occurs when learners generate an account of why something makes sense to them. They may explain why a mathematical step is valid, connect a scientific sentence to a mechanism, or identify how a new example relates to an earlier principle. The audience can be the learner alone. The important feature is the construction of a connection that was not simply copied from the instructional material.

Research by Chi and colleagues examined how prompting learners to explain a text could support understanding. A proposed mechanism is that explaining exposes gaps and encourages learners to integrate new information with existing knowledge. In a worked solution, a learner might move from noticing that a number changed to identifying the operation and its justification. However, producing more words is not automatically better. A fluent paraphrase can leave the underlying misconception untouched, and a confident explanation can be incorrect.

Consider a lesson about equivalent fractions. A learner sees one half rewritten as three sixths. Repeating that the fractions are equal provides little evidence of understanding. A more useful explanation is that each original half has been divided into three equal parts, making six equal parts in the whole while preserving the amount selected. The teacher asks the learner to connect a partitioned rectangle with the numerator and denominator. A second prompt asks why multiplying only the numerator would change the amount. These questions make the relevant relationship available for inspection.

Good prompts target a precise reasoning opportunity. Ask which principle licenses a step, what assumption is being made, or how a diagram and sentence describe the same event. Early in learning, model a short explanation and provide a focused question. Later, let learners identify where an explanation is needed. Requiring a paragraph after every trivial step may interrupt the task and increase burden without adding insight. Spoken, written, sketched, or supported communication can serve the same purpose when the reasoning is accessible.

Feedback is essential when learners lack enough knowledge to evaluate their own accounts. A partner or teacher can ask for evidence, offer a counterexample, and help repair a connection. An AI tutor can propose prompts, but its acceptance of a response is not proof of correctness. Educators should inspect whether it rewards genuine justification or merely familiar vocabulary. Learners can also compare their explanation with an authoritative model after attempting their own, noting what changed and why.

To evaluate the activity, look beyond the explanation itself. Ask learners to solve a related problem, distinguish a correct solution from an attractive error, or explain the concept after a delay. Transfer to a different representation is especially informative in the fractions example. Self-explanation is most useful when it directs attention to meaningful relationships and supports revision. It should not become a test of verbal confidence or a demand that every learner narrate all thinking aloud. The educational goal is a better connected understanding that learners can use independently.