Learning Science · Evaluating emerging learning tools

Immersive environments and transfer to real tasks

Evaluate a simulation by its learning task and transfer conditions rather than by the feeling of being present.

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

After reading, you can

Distinguish engagement and presence from learning.
Identify what a simulation represents and omits.
Design a transfer check outside the simulated task.

A convincing environment can still teach the wrong thing

A simulated laboratory can make invisible processes visible or allow repeated exploration. A vivid scene can also demand attention for navigation, controls and decoration. Feeling present in the environment and understanding the target mechanism are different outcomes.

Begin with the question the environment helps a learner investigate. A useful simulation lets the learner make a prediction, manipulate a relevant variable and inspect a meaningful consequence. Realism is valuable when the represented feature matters for the goal.

Every simulation has a model

A simulation chooses which variables, rules and constraints to include. A particle animation may illustrate a simplified mechanism while omitting interactions that matter in other conditions. Explain the representation and its limits before treating its outputs as evidence about the world.

A learner needs to distinguish changing a model parameter from establishing a real causal law. The simulation follows its programmed rules; observations of the real system are a separate source of evidence.

In the pendulum teaching example, length can be changed while other model conditions are held fixed.
Activity	What is observed	Scope
Change length inside the programmed model	The represented period under those rules	A consequence of the model’s assumptions
Explain a different diagram	Use of the relationship in a changed representation	Evidence about that conceptual connection
Measure a real pendulum	Real observations under actual conditions	Requires its own measurement and comparison with assumptions

Connect actions to explanations

A learner can discover which button earns points without understanding why a result occurs. Ask for a prediction before an action and an explanation afterward. Compare a case that fits the current rule with one that tests a boundary.

Provide a route that does not require unnecessary navigation skill or a particular sensory mode. Where immersion adds an access barrier, a readable diagram or desktop model may serve the same conceptual purpose more effectively for that learner.

Test the intended real-world connection

If the goal is interpreting a circuit, assess a new circuit representation outside the simulation. If the goal includes operating a real device, simulation performance alone does not establish safe practical competence. Relevant supervised assessment and domain requirements still apply.

Evaluate the actual combination of task, guidance, interface and learner. A simulation can be useful for one outcome and unhelpful for another. Novelty, enjoyment and time spent are informative experiences, but none directly substitutes for the learning outcome.

Worked example

A model of a pendulum

What can a pendulum simulation show, and what would require separate real observations?

A simplified simulation allows changes in length while holding other model conditions fixed. The learner predicts how the represented period changes, runs the model and explains the pattern. The lesson labels the assumptions rather than claiming the animation is a complete pendulum.
A later task shows a different diagram and asks which variable matters under the same assumptions. Practical measurement of a real pendulum would require its own observations and attention to departures from the model.

It can show consequences of changing length under its programmed assumptions. Applying that pattern to a changed representation tests a connection; a real pendulum measurement and practical competence require their own observations.

Try it yourself

Think through each question before opening the answer.

Why is solving a task inside a simulator insufficient evidence of transfer?
Show the answer and reasoning
The simulator may supply cues, simplify demands or reward interface knowledge. A relevant task outside that arrangement is needed.
Success may depend on interface cues or simplified rules that disappear elsewhere. A changed task is needed to test the particular intended connection, with its support conditions described.
What should accompany a visually realistic model?
Show the answer and reasoning
An explanation of represented variables, programmed assumptions, omitted features and the scope of conclusions.
Visual realism does not disclose the model’s omitted variables or assumptions. Naming those features connects the displayed behaviour to the limits of the conclusion.

What to carry forward

Presence, engagement and learning are distinct.
Make model assumptions and task goals visible.
Check transfer with a relevant changed task.

References

Sources supporting the explanations and learning checks in this lesson.