You’ve now reflected on your prototype and your idea and may have some ideas about how you could change it to make your idea better.
The fifth phase of the human-centred design process is testing: putting your design in front of real people to find out if they understand it and can use it successfully. Testing is how you evaluate your human-centred design and ensure that it meets the needs of your user.
The most common way of testing a prototype is to show it to someone and carefully observe how they interact with it. You can do this informally by asking a friend or colleague to try it and learn a lot.
A more formal approach is moderated usability testing. This helps gather hard evidence about the success of your design and can help solve disagreements based on personal taste.
Step 1: Develop your test hypothesis
- How is your design meant to help the person using it?
- What can you measure to evaluate if you’re successful?
For example, if your prototype is designed to help Tariq find his trainers more easily in order to get to work on time, you should test the task of finding trainers. You could measure the time he takes to find the right pair and if Tariq successfully gets to work on time when using the prototype.
Step 2: Plan your test
- When and where will they be?
- What equipment do you need?
- How will you record the test?
Audio, video or screen recording is easier than taking written notes. Prepare a script so you give everyone the same information, set them the same tasks and ask the same questions. You will need to give participants consent forms to sign, particularly if you’re recording, and keep their personal data securely.
Step 3: Recruit your participants
In our example we would test with Tariq, but if we were designing a solution for young mothers who work, we should test with young working mums to ensure accurate results. For most user tests, just five participants are enough to uncover most problems with a design.
Step 4: Run your test
Make sure you get informed consent and explain the context to the participant. Make it clear that you’re testing the design and not the person: any problems they have are your fault!
Encourage them to talk out loud as they complete the tasks, and watch carefully for what they do. If they have problems or don’t like something, try not to guide them or get defensive. Instead ask open questions to clarify what’s unclear, and what they might prefer.
Step 5: Evaluate your results
Review your notes and recordings and look for the top issues you’ve discovered.
- Can you summarise the top five insights from your test?
- What are our suggested solutions?
Writing a very simple one-page report will help you clarify and share your findings.
There are other techniques that allow you to test and evaluate designs:
Expert reviews are when an experienced professional steps through your design and offers feedback based on their professional expertise. They are good for spotting common design problems and usability issues.
Surveys allow you to ask the same questions of a group of people when they see your prototype and keep the responses structured for easy analysis. A survey could be as simple as a single question, but usually contain a few multiple choice questions and an opportunity for written feedback.
A/B tests are often used for digital products. Two variations of a design are tested against a criterion, for instance, finding out if a red or blue button encourages more people to select or buy an item. The variants are shown randomly to different website visitors and measured using digital analytics.
Digital analytics track and report on user behaviour anonymously, gathering information whenever someone uses a digital product or prototype. Reports show things like the route users take through a site or the places they are most likely to click on a page. The information helps you spot problems and make decisions about changes.
Have your say
- What techniques might you use to test your solution for Tariq?
- What would your test hypothesis be and how would you measure success?
Share your ideas with others in the Comments.
© University of Leeds