01 · Learn the idea
Set the criteria first
Criteria are the standards a design must meet, written before it is tested. For an insulated container, they might be that the ice stays frozen for thirty minutes, the container holds together, and it uses only the allowed materials.
Criteria written afterwards are useless, because they will be shaped by what the design happened to do.
02 · Build understanding
Test against each criterion
Go through the criteria one at a time and record how the design performed on each, with a measurement where possible.
A design usually passes some criteria and fails others, which is more useful than a single verdict of good or bad, because it says exactly what to work on.
criteria set first → test each → record → improve the weakest
03 · Build understanding
Trade-offs
Improving one thing often costs another. Thicker insulation keeps ice frozen longer and makes the container heavier and bigger.
Recognising a trade-off is part of evaluating honestly. Say what was gained, what was given up, and why that exchange was worth making.
04 · Build understanding
Evaluation feeds the next version
The point of evaluating is to decide what to change. Pick the criterion the design failed most clearly and change the part responsible.
Change one thing at a time, for the same reason as in an investigation: if two changes are made at once, you cannot tell which one helped.
Apply and explain
Your turn
Explain why criteria must be written before testing. Give three criteria for an insulated container. Explain why testing against each criterion is more useful than one overall verdict. Then explain what a trade-off is and why only one change is made between versions.
Use paper or explain your thinking aloud before opening the solution.
Check your reasoning
Criteria must be written before testing because criteria written afterwards will be shaped by whatever the design happened to do, so they cannot judge it honestly. Three criteria for an insulated container: the ice stays frozen for at least thirty minutes, the container holds together and does not leak, and it uses only the allowed materials. Testing against each criterion is more useful than a single verdict because a design usually passes some and fails others, and knowing which is which says exactly what to work on, while "good" or "bad" says nothing actionable. A trade-off is when improving one thing costs another, as thicker insulation keeps the ice frozen longer but makes the container heavier and bigger. Evaluating honestly means saying what was gained, what was given up and why the exchange was worth making. Only one change is made between versions for the same reason as in an investigation: if two changes are made at once and the design improves, you cannot tell which change was responsible.
- The reason criteria precede testing.
- Three specific, testable criteria.
- Trade-off defined with an example, and the one-change rule justified.
Ready to practice?
Try the questions, get feedback, and review your reasoning.
Data and explanations practice →