Learning Science · What does a learning measure measure?

Retention, understanding and transfer as outcomes

Choose outcomes that distinguish keeping knowledge, explaining it and applying it to a changed problem.

Use the examples to connect this lesson with the earlier topics. No specialist equipment is needed.

After reading, you can

Design a delayed check for retention.
Distinguish recall from explanation and transfer.
Keep outcomes separate when interpreting an intervention.

Choose the outcome before the method

Two learners can score equally on immediate recall and differ in their later explanations. A study method is therefore not simply “effective” without specifying an outcome. Retention asks whether relevant learning remains available after time passes. Understanding concerns relationships and explanations. Transfer asks whether learning helps with a task that differs in relevant ways from the practice.

These categories can overlap, but they are not interchangeable. An explanation requires remembered knowledge, and a transfer task may require an explanation. The assessment must reveal which contribution is being claimed.

Make a delayed check interpretable

Suppose a learner studies ten words and recalls them immediately. A later recall test adds evidence about retention across the stated interval. Record whether the learner practised in between, received cues or saw the answers again. A delayed result reflects the learning history that actually occurred.

A return test can itself change learning. If two groups receive different numbers of tests, their later difference may partly reflect that extra practice. Research designs should make these opportunities explicit.

Change the right features

A learner practises proportional pricing with apples. Replacing apples with pears while preserving an obvious price-per-kilogram cue is a modest change. Asking the learner to choose between a proportional and a fixed-fee model changes the decision more substantially. Describe the change rather than assuming that the word “transfer” identifies one uniform difficulty.

A novel-looking task may still reproduce every trained step. Conversely, familiar language can conceal a new relationship. Check what the learner must notice, represent and do in the new task.

Report a profile of evidence

An outcome set might include delayed recall of core terms, an explanation of a mechanism and a new application with the method left unspecified. Keep their results distinct. An average that mixes them can hide a useful gain or a serious gap.

The right set follows the purpose and available resources. A beginner lesson need not test every possible application. State the assessed interval and task changes, preserve examples of work, and identify the next untested extension.

A scale lesson can be followed by complementary outcomes.
Check	Capability sampled	Condition to record
Recall scale meaning later	Retention of the relationship	Delay, cues and intervening study
Explain proportional drawing length	Reasoning about scale	Whether key relationships are supplied
Solve a fresh route problem	Representation and application	Method cues and how the problem differs

Worked example

Three checks after a map lesson

After a map lesson, recall succeeds but a fresh route problem fails. How should three outcomes guide the next explanation?

After studying a scale drawing, a learner recalls what the scale ratio means, explains why doubling drawing length doubles represented distance and solves a new route problem without a worked template. These are complementary checks.
If recall succeeds and the route task fails, inspect the route representation and method choice. The evidence suggests a specific teaching need; it does not justify saying that nothing was learned.

Keep the recall success and inspect route representation or method choice. Recall, explanation and application supply a profile; one failure does not show that nothing was learned.

Try it yourself

Think through each question before opening the answer.

A method improves next-day recall of definitions. Does that establish improved problem solving?
Show the answer and reasoning
No. A relevant problem-solving outcome must be measured. Retention of definitions may help, but the additional claim needs evidence.
Definitions and problem solving are related but different targets. The broader claim requires a problem that samples method selection and use.
Why record intermediate study between two tests?
Show the answer and reasoning
Because the later result depends on those learning opportunities. Otherwise the interval alone can give a misleading account of what was retained from the first lesson.
The later result reflects the actual intervening practice and cues. Describing only the time interval would omit part of that learning history.

What to carry forward

Name the outcome and interval.
Describe how the application differs from practice.
Use separate outcomes to locate the next learning need.

References

Sources supporting the explanations and learning checks in this lesson.