Monday, 10 February 2014

Interview with the usability-expert Wouter Sluis-Thiescheffer, KidsRnD

The Miffy Tablet Usability Test

Remco Pijpers, CEO of Mijn Kind Online and POSCON Network Member  interviewed Wouter Sluis-Thiescheffer, KidsRnD

Can you tell me more about the Nijntje project?
The Miffy Tablet Usability Test was conducted by me in 2012 with User Intelligence. The study was setup to find out about the usability of tablet applications for very young users (<6). Miffy (Nijntje) has iconic status in the Netherlands and is internationally famous. My colleagues and me thought that an iconic character in an app will indefinitely attract attention and it was assumed it would set a high standard upon its release. 

The app landscape in 2012 was (and still is) quite rough, especially for the very young and their parents. App catalogues (like Apple’s appstore and Google’s Play) did not have a clear age indication, nor were the categories providing clear direction. For example the term “educational” was used for both “useful in the educational setting” and for “this app will teach your child something new”. Alternatives for making a choice, like review sites, were released every day, each with their own review process, target group and rating system. Making an informed choice is therefore a laboursome process. So how were parents supposed to find an appropriate app?

The release of “Miffy in the Garden” was a relief in that sense. The character Miffy is connected with a clear age indication, safe and appropriate content is guaranteed. The app was a popular download from day 1. It was also a very well timed release, the public press started to draw attention to the fact that very young children were experienced tablet users. YouTube was starting to fill up with examples of excited parents showing babies as young as 2 months old interacting with a tablet.

However a structural approach to developing guidelines for the usability of a tablet by the very young had yet to be developed. Therefore I thought it would make an interesting case to subject the first release of Miffy to a usability test with 2-4 year olds.

Can you briefly explain the app “Miffy in the Garden”?
Miffy in the Garden is an app rich in different activities. The app contains the interactive picture book “Miffy in the Garden” and three games: (1) counting carrots; (2) shadow game and (3) make your own garden.

The book can be read to you with the prerecorded voice, it can be read by the child itself and it can be read accompanied by a custom recorded voice.

With that functionality, the supports two usage scenario’s. (1) A caretaker (e.g. a parent or older child) reading the book “Miffy in the Garden” to and with a child and (2) the child using the iPad on his/her own.

So, what did you do exactly and how did you test?
The app was tested with seven children, two 2-year olds, two 3-year olds and three 4-year olds. Since I was mostly interested in the usability for young children, I only tested the second scenario, a child using the iPad on his/her own.

The two-year olds were tested in the home environment, the older children were tested in the usability lab. To keep the children comfortable in the one-on-one setting, the parent was always close by and visibly present to the child. Naturally, the parents were asked not to interfere with the activities and responses of their child.

The children were asked to perform three tasks: to find and launch the app, to play each game and to read the book.

What kind of issues did you find?
Overall, the Miffy app was very well designed. The visual language, the animations and transitions are clear and calm matching the Miffy style. The voice over (and reading voice) is recorded in a very high quality, crisp and clear, calm and comforting throughout the app. The quality of the voice is very important for this young age group as they are in the middle of their language development.

Also, all children found it easy to find and launch the app. They were all familiar with the “swipe to unlock” bar and flipped through the pages looking for the Miffy Icon to launch the app. The icon was available from a page.

However, the two and three year olds had considerable problems to go through the splash screen.
They expected that the character would take them to the next step, as they tapped on Miffy, ignoring the country flags. Although it is understandable from a business point of view (one app for all stores), two- and three year old children cannot be expected to read nor to understand country flags. They would ask the parent for help or would simply give up.

Another important issue was found in the games. The younger children found the “fruit shadow game” slightly too difficult. The two year olds and one three year old had a tendency to use multiple fingers (some from one hand and some from both hands) or use one finger and lean on the screen with the remainder of their hand. The fruits would not respond very well to the gestures of the two year olds. 

Also, the multi-finger input would sometimes lead to unwanted behavior. In some cases the app closed and in others the app changed to the previously opened app. The four year olds found the games too simple and were easily bored.

Finally what is important to mention is that when using the iPad in the home environment. For the two- and three-year olds the iPad is large and heavy device. When reading from the iPad on the couch they often need one hand to support the iPad in the right viewing angle while they interact with the other.

What changed and what did not after that first release?
In subsequent releases of the app the app improved on accounting for the motor skills for the very young. The Miffy app will not allow a swipe from the Miffy app to a previously opened app directly, allowing a two year old for example to accidentally send emails. Also, the app will not close any more after a grasping gesture. And the accuracy for controlling the games has been improved. 

The splash screen remained the same and so did the inability to recover from accidentally starting the record functionality.

So in general, how do you go about when testing with young children?
First I estimate what kind of information I need and whether the children in the target age group can provide me with this type of information. In the Miffy case, it was important to find information about the children’s behavior. Therefore it was crucial to observe the children directly. 

In other cases, when Ì want to have a conversation about motivations, or when I want to test with a functional design I consider to work with experts about the user group. Children in this age have too short a concentration span to go through an hour of usability testing. In addition, their verbal skills are not yet matured enough to reflect reliably on what they do. So, instead of children I involve for example parents, kindergarten teachers etcetera. These experts are then asked to go through the application on behalf of the children with a validated evaluation method.

When working with young children directly it is essential that they are comfortable in the testing environment. In the lab setting I often play introductory games with the children while their parents are close by. Once the child is comfortable with the test moderator, I often ask the parents to move a bit more into the background. Testing in the home environment is a good alternative. 

Children are sensitive to power-structures in the adult-child setting. This can influence a test heavily as children feel the need to “do it right” or to give “the right answer”. To soften this effect even more, I sometimes work with more children in one setting. For example in focus groups or in a peer tutoring setting. Peer tutoring puts one child in the role of a teacher, teaching a younger child what an app is about. This setting is very resourceful. The explanations of the teaching child provides insights in the mental model of the app for that child, coping strategies when stuck etc. The younger child provides insights in what is intuitive and what is not.

When working with children in a group setting there is even more to take into account, for example age differences, gender differences and group size. Homogeneous groups are recommended, both in age and gender. Children are very sensitive to age differences and a difference of two years can affect their behavior already. Similarly, children of around five to eight like to be in same-gender groups as the other gender is not appreciated. Children older than twelve years of age are often best consulted in  same-gender groups for opposite reasons. Concerning group size, I often work with five to eight children in one group. A smaller number often results in parallel interviews rather than a group, a larger group is hard to control.

Do you have design recommendations when designing for a young target audience? 
The findings of the usability study became part of our presentation Tablets and Kids. Released by User Intelligence and published on SlideShare.

Test often and test early.

The context of use is most informative about the effect for example:  
  1. motor skills proficiency on the gestures,
  2. the size of the children vs the size the tablet, and
  3. the usage scenario
  • single vs supported usage,
  • usage on a table vs usage on a couch,
  • usage as first screen vs second screen
  • usage with pawns or with fingers
  • etc.
     Provide a clear app store description

Design to empower
  1. Reinforce children’s mental models
  2. Support children’s motor skills
  3. Allow for accidental success
Design to grow 
  • Integrate complexity to address children at different stages of development
Keep the navigation structure consistent and shallow 
Mind the edges, they are often used to control the tablet, not so much the app.

Wouter Sluis-Thiescheffer (website in development)

No comments:

Post a Comment