Design Thinking
Design thinking is examined as a disciplined problem-framing and solution-finding method, not as a synonym for creativity. You should be able to run the loop, choose the right tool for each mode, and show how evidence changes the design direction.
The loop and the double diamond
- Empathise: observe and interview real users in context; capture workarounds, not opinions.
- Define: turn findings into a point-of-view and a testable problem statement. A How Might We question should be narrow enough to act on but open enough to allow alternatives.
- Ideate: diverge before converging; defer judgement; use SCAMPER, analogies and extreme users.
- Prototype: build the cheapest artefact that answers the current question.
- Test: observe behaviour against predefined success criteria, then loop back. The double diamond alternates divergence and convergence: explore the problem, define it, explore solutions, deliver. The loop is iterative, but each iteration needs a stated learning objective.
Worked example
A clinic wants to cut check-in queues. Interviews show that older patients abandon the QR-code kiosk because they cannot find the code on their phone. The define statement becomes: “Patients over 70 need to confirm arrival in under 60 seconds without using a smartphone.” Three concepts are generated: a staff tablet, an SMS link and a self-service kiosk with a printed code. A paper prototype and role-play test with six users show 5 of 6 fail the SMS route because of typing errors, while 6 of 6 complete the staff-tablet route in under 45 seconds. The team pivots to the tablet and keeps the kiosk as a secondary route. The reasoning is behavioural: the winning concept removed the step where users failed, rather than adding instructions.
Traps that cost marks
- Jumping from empathy to a solution and labelling it design thinking.
- Writing a define statement that contains the solution (“we need an app”).
- Testing a polished prototype too late, when feedback can no longer change direction.
- Running ideation and evaluation in the same session, which kills unusual options.
- Reporting opinions (“users loved it”) instead of task success, time or error rates.