Learning Science · Learning from outcomes and other people
Learning associations and predictions
Explain learned predictions while distinguishing a cue that forecasts an event from a cause that produces it.
Use the examples to connect this lesson with the earlier topics. No specialist equipment is needed.
Back to the chapterAfter reading, you can
- Describe a cue-outcome association
- Separate prediction from causation
A familiar sound before an event
A particular chime regularly precedes the start of a community meeting. After experience, hearing it can bring an expectation that the meeting is about to begin. An association links aspects of experience so that one can help predict another.
In classical conditioning, a previously neutral cue can come to elicit a learned response through its relation to a significant event. The details matter: timing, predictive information and previous learning influence what is acquired. Simply seeing two things together once does not establish a universal rule about future behaviour.
Prediction is not causation
A dark cloud may predict rain without causing rain by being seen. Likewise, a chime can signal a meeting without producing the meeting’s content. A learned expectation can be useful even when the learner has no complete causal explanation.
New outcomes can challenge an expectation. If a cue appears without the expected event, learning may change. This change need not erase the earlier association; a response can sometimes return with a different context or after time. Describe the conditions instead of assuming an old expectation has vanished permanently.
Use the mechanism thoughtfully
Reliable cues can make a setting easier to navigate: a clear signal indicates when to start or where to seek help. Ambiguous cues can create mistaken expectations. In education, a signal can organise action, but it does not by itself teach the reasons behind a concept. Conceptual explanation and cue learning can contribute differently to the same activity.
Prediction also involves noticing what happens when the cue is absent. If the meeting starts just as often without the chime, the chime carries less information than expected. This illustrates why association is not only a count of co-occurrences: the surrounding pattern matters to interpreting a useful prediction.
Worked example
A website’s confirmation signal
A green banner used to accompany successful uploads. It now appears even when the intended file is missing. How should the learner interpret and verify it?
- Separate the signal from the outcome. The banner is a visible cue; the relevant outcome is the intended file being present and available. Its colour does not make the upload occur.
- Compare the two cases: a present file and a missing file can both be followed by the same banner. The cue no longer distinguishes those outcomes reliably in this interface.
- Check the intended filename and the actual upload status using the application’s dependable record. Update the verification routine rather than repeatedly relying on the old colour expectation.
The learned expectation was understandable under the earlier pattern, but the changed cue now needs an independent outcome check. A useful prediction must stay connected to the event it is meant to signal.
When one cue fits two different outcomes
Try it yourself
Think through each question before opening the answer.
Does a cue that predicts an event necessarily cause it?
Show the answer and reasoning
No. A predictive relationship may arise from common causes or regular sequences without the cue producing the event.
A chime can tell people that a meeting is about to begin because both belong to a schedule. That predictive relationship does not establish that hearing the chime produces the meeting.
Why may an old expectation reappear after a change of setting?
Show the answer and reasoning
New learning may depend partly on context while the earlier association remains available. A changed setting can alter which expectation guides the response.
A response declining in one setting does not prove that the earlier association was erased. Context and time can affect which learning guides the response, so describe the actual conditions of its return.
The meeting starts just as often without the chime as with it. Why does counting only chime–meeting pairings give an incomplete picture?
Show the answer and reasoning
It ignores how often the meeting occurs when the cue is absent.
Frequent co-occurrence can reflect a frequent event rather than an informative signal. The comparison with cue-absent occasions changes what the chime predicts.
What to carry forward
- Learned cues can guide expectations.
- Prediction and causation differ.
- Check whether a cue still tracks its intended outcome.
References
Sources supporting the explanations and learning checks in this lesson.