User testing with children
working with children and the key points that user researchers need to remember when work
Fowlam’s CEO, Frances Brown, is a trained primary teacher and taught for nearly five years. She worked with children with special needs and also carried out complex psychological research with children with Down syndrome, autism, specific language impairment and dyslexia. In her present role with Fowlam, she has put the skills she developed both as a teacher and a researcher to use in testing products with children for companies such as LEGO. Here she talks about her experiences of working with children and the key points that user researchers need to remember when working with younger participants:
Testing with children requires a very particular approach – it’s not enough just ‘dumb down’ a process that’s used with adults. The age of the children is key – different age groups will have very different attention spans and levels of comprehension.
It’s very important that both the people designing the study and the facilitators have experience of working with children, because testing with children requires an understanding of how to interact with them and how to interpret their responses. Children tend to be much less patient and forgiving than adults – if the study is too long, too difficult, lacking in energy or badly organised, the child may shut down, get upset, get distracted or ask to leave. Once the child gets to that point, it’s almost impossible to salvage the situation.
Even if the children are engaged and responsive, they won’t always give you great data. In my experience, children will often answer a question even if they have no idea what they’re being asked. They rarely say that they don’t understand and they will sometimes claim to understand something (or genuinely believe they understand something) even when they don’t. If you don’t establish a child’s level of comprehension early on, you may end up in a very frustrating situation where you’ve happily gone through the entire test scenario only to find right at the end that the child simply cannot complete the task and all of their previous data is useless.
The testing scenario can be graded, so that different age groups are given different tasks and questions, to ensure that they are pitched at the right level. But the downside of age-based cut-offs is that they can be quite rigid -they predetermine the level of testing before you have a chance to really see what the children can do.
One very effective method that I use to fit the testing to the child’s ability is the threshold system. It’s a system that requires very careful planning of the research study so that at key points the child’s comprehension of various aspects of the task are assessed. For example, if a child says they understand a process, the tester might ask them to act the process out or to show a doll how to do it, to ensure that they actually do know how to complete the process without adult support. This method ensures that you can be confident that the children only answered questions when they really understood what was going on, while at the same time allowing scope for children to complete the most difficult tasks, if they’re able to. Being able to identify those key points, and the right questions to ask to ensure comprehension without making the testing session overly long or complicated takes quite a lot of skill and experience.
Keeping children on task is another aspect of testing the requires skill. Younger children, in particular, can get very easily distracted and start chatting or playing with materials. The facilitator needs to be friendly and accommodating while at the same time ensuring that distractions are kept to a minimum. The facilitator also needs to be organised and prepared so that the testing session keeps moving along – any downtime can mean a child switches off. It takes a lot of energy to ensure that testing sessions are swift, organised and accurate – it’s a real juggling act!
There are other, practical considerations that need to be taken into account, such as when and where the testing is carried out. Younger children can be very tired in the late afternoon, especially if they’ve been at nursery or school that day. Adults tend not to like morning testing, but many children are at their sharpest before 11am. If parents and siblings come along then you need to provide drinks and a comfortable, safe place for them to wait. And remember, if you’re giving an incentive to a child, you must provide something for their siblings too (even if it’s something smaller, like stickers.) Parents will not appreciate having to drag a crying child away because their sister got something they didn’t!
Difficult as it is to imagine how adult users will engage with a product, it’s almost impossible to guess how children will react. For example, a child might be able to download and install and app with no difficulty, but then be totally unable to follow the basic instructions to get the app started. Hands-on user testing is absolutely essential to get a clear understanding of the areas where you need to support children’s abilities in order to ensure they can engage successfully with a product and really enjoy it.
I absolutely love testing with children – it’s challenging, but great fun. I’m happy to talk to companies who want to test with children about how they can get the most out of their research.’
Contact Frances, email@example.com if you would like any advice about carrying out user research with children or if you’d like Fowlam to plan and implement a user research study for you.