Learning Science · Learning with digital environments

Simulations, models and their assumptions

Use a simulation to test relationships within a model while checking what the model leaves out.

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

After reading, you can

Identify a simulation’s variables and assumptions.
Distinguish simulated results from evidence about the world.

An adjustable world

A simulation lets learners change the mass of an object sliding down a ramp. They predict whether travel time changes, run the model and compare results. The activity can make a relationship easier to inspect, but the answer depends on how the model represents friction and other conditions. The screen is showing a model’s consequences.

Predict, vary, explain

Name the question and the relevant variables before manipulating controls. Change one factor deliberately when the goal is isolating its role, record the result and explain it through the model. Random exploration may reveal features, but it can make causal interpretation difficult when several inputs change together.

Ask what is fixed, omitted or idealised. A frictionless model can teach a useful limiting case, while failing to predict the exact motion of a real object on a rough surface. A correct simulation result does not independently prove that its assumptions describe the target system. Compare with suitable real evidence where the claim concerns the world.

For an ideal ramp comparison, first specify the model and what changes.
Role	Specified feature	Why it matters
Varied input	Mass of the sliding object	The intended comparison
Fixed conditions	Ramp angle, start position and initial motion	Other changes could alter travel time
Idealisation	No friction or drag in this limiting case	Real surfaces can violate the assumption
Further world evidence	Measurements and conditions on the real ramp	A simulated output does not validate its own assumptions

Match the simulation to the skill

Simulations can reveal otherwise invisible quantities, offer repeatable cases and allow safe exploration of a conceptual model. They may not develop physical equipment handling, tactile judgement or all sources of measurement uncertainty. PhET’s research overview explicitly distinguishes conceptual goals from laboratory skills that simulations may not address.

Make the controls, representations and feedback accessible. If operating the interface is difficult, teach it separately or provide another route. After exploration, check the target on a new representation or real task; familiarity with the simulator’s controls is a narrower outcome.

Worked example

Testing the ramp model

In an ideal frictionless ramp model, what must stay fixed when asking whether changing mass changes travel time?

The learner holds ramp angle and starting position fixed, changes mass and records simulated time. The model gives the same time in its frictionless setting. The learner explains that this conclusion belongs to the specified model, then predicts what an added drag force might change. A later diagram question removes the controls and asks for the relationship. A real experiment would require checking measurement and physical conditions rather than assuming the simulation already settled them.
The learner can also ask which quantities the model displays directly and which must be inferred. A simulation that shows an acceleration arrow supplies information that a real observer may need to derive from position over time. Removing or changing that display in a later task reveals an additional transfer demand and helps interpret differences in performance.

Keep the ramp, starting position and initial motion fixed, and state the frictionless sliding model. Its output follows its assumptions; a real ramp or equipment-handling claim requires further evidence.

Try it yourself

Think through each question before opening the answer.

Why record the model’s friction setting?
Show the answer and reasoning
It is a condition that can affect the predicted relationship. Without it, the result may be applied beyond the model’s scope.
The friction setting is part of the model, not a decorative control. Applying its output to a rough surface without checking conditions changes the claim.
Does successful virtual equipment use prove real handling skill?
Show the answer and reasoning
No. Physical setup, tactile information and practical errors may be absent. Those skills require appropriate practice and checks.
A simulator can omit setup, tactile cues and physical error sources. Successful control of its interface demonstrates a narrower capability than handling real equipment.

What to carry forward

Inspect assumptions and manipulate variables deliberately.
Check transfer beyond the simulation’s interface and model.

References

Sources supporting the explanations and learning checks in this lesson.