DT
IB Design Technology
SL · Lesson 33 · Model → Test → Refine
01 / 14
Lesson 33 · Unit 4

Model → Test → Refine

Use drawings, physical prototypes and CAD models as instruments for gathering evidence—not as display pieces.

80 minutesB2.2.6
Guiding question: How can prototypes generate evidence that changes a design for the better?01
Today’s destination

By the end of class, you should be able to…

01

Choose the right model

Match the representation to the design question.

02

Gather evidence

Distinguish measurable data from user/client feedback.

03

Analyse results

Turn evidence into a design judgement.

04

Refine iteratively

Make a targeted change and test again.

Success check: Explain the concept accurately, then use it to make a justified design decision.
B2.2.6 · prototyping as evidence02
?
B2.2.6

A prototype is a question you can interact with.

The best prototype is not always the most realistic one. It is the one that helps answer the next important uncertainty.

1
“Will the user reach this control?”
2
“Does the mechanism jam under load?”
3
“Can a manufacturer interpret the assembly?”
4
“Which form does the intended user prefer—and why?”
Prototype purpose starts with a question03

Drawing

Best for proportions, component relationships, dimensions, assembly communication and quick feedback.

Physical prototype

Best for touch, reach, grip, space, motion, fit, perceived comfort and real-world interaction.

CAD model

Best for geometry, variants, interference, visualization, rapid-prototyping preparation and some forms of simulation.

B2.2.6: designers select among drawings, physical prototypes and CAD models according to the evidence they need.
Select the tool that fits the uncertainty04
Data

Measured / observed

1
Time to complete a task
2
Reach distance
3
Number of errors
4
Force / deflection
5
Dimensions or clearances
+
Feedback

Interpreted by people

1
Preference
2
Perceived comfort
3
Confidence
4
Confusion / frustration
5
Client priorities
Strong development often uses both: performance evidence + human response.
Gather relevant data and feedback05
01

Model

Represent the uncertain feature.

02

Test

Use a planned method with users/clients or measurements.

03

Analyse

What does the evidence mean?

04

Refine

Change the feature for a reason.

05

Retest

Check whether the change improved the result.

Model → test → analyse → refine → retest06

Example: lid opening force

Question

Can the intended user open the lid with one hand without excessive force?

Method

Test three hinge geometries with the same user task and record success/errors + user confidence.

Decision rule

Keep a geometry only if the user completes the task reliably and rates control as acceptable.

Refinement

Change hinge leverage / grip area—not random aesthetics.

Define what evidence will change the design07
Choose the best prototype

You need to test whether a handle is comfortable during a 30-second carrying task.

B2.2.6 · representation must match the question08
User-centred evidence

Do not ask only “Do you like it?”

Observe behaviour

Where do they hesitate? Re-grip? Make errors? Compensate? Ignore a feature?

Ask targeted questions

What felt difficult? Which option gave more control? What would stop you using it?

Keep task/context realistic

A prototype tested in an unrealistic task can produce misleading evidence.

Record the result

Photos, measurements, short quotes, tables and annotated changes can make iteration visible.

Gather evidence from actual potential users / clients09
Evidence → inference → change
Result4/5 users miss the rear control.
InterpretControl visibility/reach is poor.
LinkConflicts with usability specification.
ChangeMove control to front edge + increase contrast.
RetestRepeat task under same conditions.
Do not skip the analysis step10

Prototype theatre

Building something impressive that answers no defined design question.

Confirmation bias

Testing only to prove the idea works rather than discover weaknesses.

Uncontrolled comparison

Changing many variables at once, so the cause of improvement is unclear.

No retest

Making a refinement but never checking whether it actually helped.

Better prototyping = better questions + better evidence11
Retrieve before you reveal

Can you explain the chain without notes?

Why might a low-fidelity model be better than a high-fidelity model early in development?
It can isolate a design uncertainty quickly and cheaply, making iteration faster without spending effort on irrelevant realism.
What is the difference between “feedback” and “data”?
Data are measured or observed results; feedback is a person’s interpreted response, preference or insight. Both can guide development.
Why does B2.2.6 include drawings, physical prototypes and CAD models?
Different representations generate different kinds of evidence; designers must select the representation appropriate to the question.
Spiral retrieval · A2.2 + B2.212
Tomorrow’s IA move

Use this loop to decide between your top concepts.

Criterion C asks you to evaluate redesign ideas against the design specifications through testing and user feedback.

Prepare tonight: choose 2–4 high-value specifications and one piece of user feedback that can actually discriminate between your ideas.
Bridge to IA Criterion C evaluation13
Exit ticket

Before you leave…

1. Name one uncertainty in your redesign that needs testing.
2. Choose the most appropriate model type for that uncertainty and justify it.
3. State what data or feedback would cause you to refine the idea.
Next lesson

IA Day 10 — Evaluate Concepts with Users & Specifications

Keep the evidence chain moving: user → research → decision → test.

Curriculum alignment
B2.2.6: prototypes gather data and feedback from potential users and clients; select and use appropriate drawings, physical prototypes and CAD models to gather relevant data and feedback, then analyse and develop the design iteratively.
Lesson complete14 slides
Designed by David Xu · © 2026 David Xu. All rights reserved.Educators and students are welcome to use these materials for non-commercial teaching and learning with attribution. Please share the original link when possible. Reposting, redistributing modified copies, removing attribution, or commercial use requires prior permission.
IB Design Technology SL · Lesson 3314