
plan a fair test of one change to our machine, carry it out, and use the results to make the machine work better.
Phase 4: Build
Do Now. Priya’s team changed three things at once: a taller ramp, closer dominoes and a new car. The machine worked better.
Which change made the machine work better?
Junior engineers, today we revise. To revise means to change a design to make it better. Good engineers revise with a plan, not by guessing.
A fair test plan has four parts. 1. The ONE thing I will change. 2. What I will measure or count. 3. What I will keep the same. 4. How many trials I will run. Use the same tools each time, like the same ruler and the same marble.
Worked example 1: Malik’s plan. Malik’s machine stopped between the last domino and the car. Change: move the car 2 cm closer. Measure: does the car reach the bell? Keep the same: same dominoes, same car, same ramp height. Trials: 3 before the change and 3 after.
| Trial 1 | Trial 2 | Trial 3 | Bell rang (out of 3) | |
|---|---|---|---|---|
| Before the change | ||||
| After the change |
Worked example 2: reading the results. Before the change, Malik’s bell rang 1 time out of 3. After the change, it rang 3 times out of 3. The data shows the change helped. These are sample numbers. Your data will be your own.
Test in different ways. A solution should work more than once and in more than one setting. Try your fix on a different table, or with a different team member letting go of the marble. Does it still work?
Write your fair test plan with your team, then check it.
Run your plan and record the results in the table.
What did you change, and what happened? Use your before and after data to say if the machine works better.
Where did your idea for this fix come from: a test, the peer critique, or something you read or learned? How did talking with others help?
One fix at a time is how real machines get better. Next time, we see how real engineers fix a big energy problem in cars and buses.
Exit ticket: name one thing you kept the same in your fair test. Why did it need to stay the same?
How confident are you that you can I can test my device and use what I learn to make it work better.
Materials: each team’s prototype, rulers or meter sticks, the before/after table, extra tape and dominoes.
Setup: approve each team’s plan before they test. Make sure each plan names one change only.
Listen for: teams changing several things “to save time”; counting only successful trials; deciding a fix worked after one try.
Support / stretch: give a plan frame with the four parts. Stretch: test the fix in a second setting, such as a different table.