We were worried that the children would be stressed to test the prototype. They had to follow instructions and people they didn’t know would be watching.
During the introduction, present yourself as part of a team. Explain them who is behind this and try to let them relate to your position. In this case it was a well known tv channel for kids.
We emphasised that the app or product was something new, that wasn't available yet and that they were the first to see it. This had the effect we wanted, as almost all children were eager to get the test going.
Let the parents/teachers/… decide who would come in when for the test. So in the beginning, only a few children tested the app. They got the other children curious to test as well. In the end, try to get all the children who are aware of the test, to have a go at it and play around with the prototypes. This is to avoid jealousy or misconception about how useful their input is.
We absolutely wanted to avoid any feeling of lab-setting, this is how:
To put the children at ease with the whole testing situation, we kept things as comfortable as possible. We didn’t want them to have to move to unknown territory. The test took place in the lunch room of their school. The moderator sat next to the children, and was dedicated to asking and answering the questions and observing. The test observer sat behind the children, observing and taking notes. Try to make sure there is no "Us against Them" setting.
Use the day-to-day furniture the children have, for example at school. Small tables and chairs allow you to take place next to the participants and have a view on things from their position.
Instead of forcing them to sit on bigger chairs at bigger tables, place yourself on their height.
Before the actual test starts you always, regardless of the audience, make sure the participants know what they can expect and what our expectations are.
We are there to test the concept and the prototype, not the participants; there is no right or wrong answer; anything you do with the prototype is useful for us.
These elements couldn’t be stated enough regarding the age of our participants.
In all types of user tests, getting the participant to speak up their mind is hard. With children, this was even more challenging.
Having children express what they see, think and do isn’t easy for a complete stranger. Therefore it’s really helpful to allow them to do the test with a peer they know. This way they have a companion who can help them, who can talk to them or ask them questions and who can discuss the different topics with them. For us this meant valuable data, as these mostly come to the surface when people disagree on a topic!
Often user tests are completed by following a scenario. But this app is all discovery, so we opted for a real-life situation to do our test.
Allow children to discover whatever you have prepared for them. They need more time to process all the information and to make up their mind. Only start asking questions after they have seen most of the content you had prepared.
When the setup is right, the children are ready and talkative, a few more elements can mess up your good work. You need to start with good data to end up with good data.
Work with a super high fidelity prototype. It is very hard to explain to children that a prototype isn’t something completely functional, that the content isn’t correct yet or that the design isn’t finished yet. You have to be able to show them a product that could have been live already. Work with actual, new ‘content of the day’ and put everything inside a design that is as good as ready.
To make sure the participants have enough time and input to think about their answer, we almost constantly rephrased their answers and our questions. This may seem awkward as a test leader in the beginning, but soon enough you’ll find out that discussing an answer leads to deeper insights. As always, it is very important not to be suggestive or judging in the way you ask your questions!
If you want to be able to use your gained insights to a wider population, make sure you test enough participants from a homogenous group. For children, this starts with the age; gather enough data per age-category. You won’t have the same results with a 6 year old or a 9 year old!