Photo by Hal Gatewood on Unsplash

Doing Usability Testing with Children

Bianca Copello
Nov 4, 2018 · 7 min read

So you’re designing a product for kids. Great! Kids are fun, clever, wonderful little blobs of joy. They’re also unpredictable and unreliable. There are a number of things to consider when testing children that can make the task seem daunting, but it can be a rewarding experience and give you opportunities to think creatively.

For the purposes of this article, I’ll be referencing children under twelve years old. It’s also important to note that this article is written based on my own experience with children, but I have a long way to go before I can say I have any expertise in testing children, and I am not a parent myself.

Challenges

Children have an internal set of logic that often varies wildly from an adult’s. This can make their responses difficult to predict and direct towards helpful answers. They also have very limited life experience, which can render some types of questions useless. For example, a child’s responses about frequency or length of time are going to be nearly always inaccurate because a child’s sense of time is not well-developed. Something that happened a week ago could very well seem like a year ago to them. Likewise, quantities are also problematic, as a “ton of money” for an eight-year-old might refer to their $5 allowance.

Some children are very vocal and will have no problem engaging you and answering your questions — sometimes with maybe too much detail. Other children, especially the younger ones, may have a hard time opening up and may not have the verbal skills to give complete answers. You will have to be adaptable and sensitive to the child’s mood and abilities.

The answers children give you often have to be taken with a grain of salt. Most children are not deliberately trying to lie, but they might believe things that are inaccurate and speak very confidently about them, which makes it hard to know what is real and what is not. Occasionally, there are very clever children who “lie” as a joke — this is not out of malice, but to see our reactions. They get delight out of tricking an adult and seeing the reaction.

It cannot be understated how critical maturity is to all and any interactions with children. Maturity is closely tied to age, but it can vary wildly by child, so you must gauge it by interacting with them. The younger or more immature the child is, the harder it will be to get actionable data from them. Their age and maturity will determine whether you let them read materials on their own, or to what extent you have to break down questions and simplify testing. There are many guides out there that can express skills children generally have at each age level, but it’s best to always underestimate those abilities until proven otherwise. For example, don’t assume that because a child is of reading age that it means they are a fluent reader and can handle testing a game with full sentences or many words.

These are only some of the challenges involved with children. If you are planning on publishing your research and are part of an academic institution, you also will face higher scrutiny and will have extra steps to take in regards to getting approval from ethical committees such as institutional IRBs.

Tips

Parental Presence: For children younger than eight or shy children, having at least one of the parents present can offer a sense of comfort and can help correct exaggerations or supply missing or confounding information during interviews. For older or very confident children, you can ask them if they need their parent present. If they say they don’t and the parents are also comfortable with it, take the opportunity to do a solo interview with them — you might get sets of answers unbiased by parental presence. At the same time, remember that parents usually know their children best and defer to their judgment.

Language: Simplify your questions and language as much as possible. For children under five or six, you may have to ask yes or no questions to get definitive answers. Be gentle, prompt and give examples if needed, and encourage children by showing interest and asking direct follow-up questions. A simple “Why?” after a yes or no question will give you better data than a rating scale. Props such as dolls might help to redirect the question by getting the child to role-play with the doll. You may have to keep them on track and bring them back to the main question. Repeat it if necessary in case they’ve forgotten it: “But what about when you do this?”

Rating Scales: Rating scales require a certain level of awareness of self and others and are often too complex for children. However, if it fits your question, visual scales might get better results. For example, a series of smileys from a crying face to a laughing face.

Expect Family: When you are testing a child, you have to be sure to include the parent. Not only can they corroborate the child’s answers and offer clarification, but their comfort and understanding of the study are as important as the child’s since they will be the ones to determine whether to allow their child to participate in your study. Along those lines, expect siblings and come prepared with ways to entertain siblings while you are testing. Not all parents have the luxury of having caretakers for the siblings, and the interactions might prove valuable to your study.

Connect with Them: Get down on their level so you can see them eye to eye. Talk to them honestly, and don’t be personally offended by anything they say or how they react to you. Don’t try to force their honesty or engagement, give them time and space to warm up to you if they need it. Prompt them with questions about their favorite toys, books, or TV shows. Bring a focus item such as a toy to gain their interest and let them play with it while you interview them. Complimenting them on something special they own can open up the conversation as well. Remember they are a little person with distinct wants, needs, and fears, even if they’re still under formation. They deserve your respect and attention, just as any adult participant does. At the same time, they also require a lot of leeway, patience, and understanding.

Compensation: It’s easy to give children stickers, candy, or even cheap toys. Ensure you have the parent’s approval and consider your choices. If it’s candy, make sure you have something small that won’t ruin a kid’s appetite, and preferably more than one type of candy in case the child doesn’t like the single option — they might feel devastated at the idea of getting a reward and then it being a type of candy they can’t eat or don’t like. If you’re getting a toy or trinket, make sure it’s not something that could be problematic for the parents — things that make a mess or make noises are not likely to be appreciated by the parents. If there are siblings present, make sure to give them something as well, and preferably the same item to avoid any fights or jealousy. Older siblings may be understanding that the sibling that did the test may get a bigger reward, but younger siblings will have a hard time with it.

Ethics: Be aware that if a child reveals an abusive situation at home, you may be required to report it to the authorities. Those situations are beyond the scope of this article, but it might be a good idea to consult with your institutional review boards or local laws to be aware of what to do if you find yourself in that situation.

Timing: Testing with children takes extra time, but at the same time you can’t expect the continued focus or engagement of an adult. A ten-minute activity may be more than their attention span. Your steps have to be very simple and very short. Your questions have to be short. But you may spend a surprising amount of time directing their efforts and prompting for responses to get successful data. Plan extra time for usability testing but also don’t be surprised if interviews are shorter than you expected.

Tactics: Ask kids to explain something back to you in their own words. This way you can test their understanding of the system. It’s important that you, or the system, has properly explained it to them first and that they’ve had some time to internalize or practice the concept. Your ways of measuring their engagement or interest may have to be indirect since children are even more likely than adults to reply affirmatively to questions. This is your chance to get creative. For example, you could test their emotional engagement with the system by giving them cutouts of characters from the system. Observing how they react to the cutouts might give you indirect clues about whether they will be able to connect to the characters when they’re in the system.

Overall, getting specific, nuanced answers from children is difficult, but testing with them can reveal serious flaws and overarching trends that can prove invaluable. A system we tested once revealed the need for incorporating sibling interactions. More importantly, it revealed that our system was too wordy for young children who couldn’t read fluently yet but too boring for the older children that could. Doing testing with children will help you get a deeper understanding of your system and is sure to be memorable. Enjoy the interactions and appreciate a peek into the fascinating mind of a child.

More information can be found at https://www.usability.gov/get-involved/blog/2015/02/working-with-kids-and-teens.html

Cover photo by Hal Gatewood on Unsplash

Georgia Tech MS-HCI

Writings from the HCI Master's students at Georgia Tech

Bianca Copello

Written by

Georgia Tech MS-HCI

Writings from the HCI Master's students at Georgia Tech

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade