Toy Cars on Three Surfaces: When the Average Hides a Problem
A fourth-grade class rolls identical toy cars down a 30 cm ramp onto three surfaces and measures how far each car travels. Mean distances: smooth tile 142 cm, short carpet 61 cm, playground grass 33 cm. Problems surface during the investigation. In one group, a different student releases the car each trial. In another, distance is measured from the ramp base to the front of the car in one trial and to the back in the next. The tile team's three distances are 140, 139, and 220 cm, and their car was swapped after a wheel fell off. The teacher has one class period left and wants the class to make a defensible claim about surface roughness.
1. Which parts of the class's data support a claim about surface roughness, and which parts must be fixed or discarded before the claim can be made?
Read the worked answer
What is supported
The overall ordering - tile far, carpet nearer, grass nearest - is large and consistent in direction, well beyond plausible measurement noise of a centimeter or two. That pattern supports the general claim that rougher surfaces reduce travel distance. The magnitude of the gaps (142 vs 61 vs 33 cm) is what makes the claim robust even with procedural flaws.
What must be fixed
- Different releasers: release height and push vary between trials and between groups, so distance differences may reflect release, not surface. Fix: one release method, such as a gate or a held ruler, used by everyone.
- Inconsistent measuring points: front-of-car and back-of-car differ by the car's length, which is comparable to some real differences. Fix: pick one reference point and state it.
- The 220 cm trial: it is far outside the other two tile runs and coincided with a car swap after a wheel failure. It should be flagged, investigated, and excluded from the tile mean, with the reason recorded. Excluding it is not cheating when the reason is documented.
- The switched car: a replacement car of different mass or wheel condition breaks the 'identical cars' assumption, so the tile mean should be recomputed from the original car's valid trials, or the trial repeated.
Claim to make now: rougher surfaces reduce travel distance under our conditions; with better release and measuring controls, we could estimate how much.
2. Explain, in energy terms, why the grass stops the car soonest, and how a student's idea that 'the grass pushes back harder so the car runs out of push' should be corrected.
Read the worked answer
The energy account
The car starts with gravitational potential energy from being raised up the ramp; that becomes kinetic energy at the bottom. The distance traveled depends on how quickly that kinetic energy is converted into other forms. Friction between the car and the surface converts kinetic energy into thermal energy in the car, wheels, and surface. Rougher surfaces - grass, then carpet - do more of this conversion per centimeter, so the car stops in a shorter distance. All three cars start with roughly the same energy, so the one that stops soonest must be losing energy fastest.
Correcting the 'push' idea
The phrase 'the car runs out of push' implies a stored force that gets used up. There is no forward push once the car leaves the ramp; it coasts. Friction is a backward force acting the whole time, and it is larger on rough surfaces. The car does not lose its push - it loses its motion energy to friction, and the rougher the surface, the faster that transfer happens.
Useful check question: if the surface were perfectly frictionless, what would the distance data look like, and why can that never happen in the classroom?
3. Design the fairest version of this investigation the class could run in one period, and identify one factor that would still be uncontrolled.
Read the worked answer
Fair-test design
- Same car for all trials; no swapping mid-investigation. If a wheel fails, note it and restart that surface's trials.
- Same ramp, same height, clamped so it cannot shift between trials.
- One release method: a ruler held as a gate, released without a push, same student or same technique every trial.
- One measuring reference: distance from the ramp base to the car's rear axle, marked with tape on the floor.
- Five trials per surface, rotating the order of surfaces so that warming or flattening the carpet over time does not favor one condition.
- Record room conditions and any obvious surface changes, such as grass being flattened by repeated runs.
Analysis
Report the mean and the spread. If the five tile runs are 138–143 cm but the five grass runs are 25–45 cm, the difference is convincing even without statistics. Show the spread on a simple bar graph with range marks.
Still uncontrolled
Grass and carpet are not standardized materials: a patch of playground grass varies in length, moisture, and compaction, and repeated runs flatten it. The class should state this limitation rather than claim the comparison is perfect, and should describe how a controlled rough surface - sandpaper of two grits, for example - would test the same idea more precisely.