Making Apps for Kids: Part 3

Some lessons learned in the making of Quick Math Jr.

Over the past few weeks we have been writing a series of blog posts about the many, MANY lessons we are learning as we create our app, Quick Math Jr. Check out parts 1 and 2, or read on for part 3!

Getting schooled

We do a lot of our user testing in Kindergarten classrooms. Testing with group of kids at a time can be chaotic, but it also has several advantages.

For us, it allows testing with hundreds of kids over the course of development. For our testers, it offers a familiar and comfortable environment, surrounded by people they know (plus one or two visitors!).

The greatest advantage, however, is the opportunity for us to collect honest opinions from our testers. When kids are playing an app alongside their friends, they talk to each other as they play. In doing so they naturally verbalise their thoughts about the app, and we are right there to hear it!

Kids this age are becoming much more socially aware. They are also developing an understanding of their own learning and starting to compare themselves and their performance to that of their peers. Much of what we typically hear in a testing session is reflective of these developmental changes.

For example…

“This is SOOO easy!”

As soon as one kid says this, you know you’re about to hear a chorus of agreement (even from that kid who saying it was too hard just moments before!). Kids this age are conscious of being ‘big kids’ now, and distinguish themselves from younger kids by rejecting things they regard as ‘babyish’. Naturally, they don’t want any of their peers to think of them as babies either, so are quick to join the “SO easy” bandwagon!

In her book, Design for Kids, Debra Levin Gelman suggests that it’ s hard to pin down exactly what makes something “babyish” but that typically “children call something ‘babyish’ when it’s not difficult or challenging enough for them”.

On the flip side, there is:

“Grrr, I HATE this game!”

This is not exactly encouraging to hear, but it is important feedback. Typically we find that this too is related to the challenge level of the app, but in this case it often means the app is too hard. While a too-easy game may be boring, too-hard is frustrating. Kids this age are conscious that they are struggling, become frustrated with themselves and project that frustration onto the app.

Takeaway: It’s hard but necessary to keep kids this age in the difficulty sweet spot. Quick Math Jr. has an adaptive difficulty engine that tries to keep players in their zone of proximal development . This means that new players are presented with easy challenges to start, then the difficulty increases based on their performance. Classroom testing sessions let us know when this is not working quite quickly enough!

“Omigosh, Anna is up to one hundred!!”

Five year olds get really excited when they see numbers over 100. It’s amusing, but it’s also reflective of their general thirst for knowledge about the world, and their developing awareness of their learning achievements.

One mistake we made in the original version of Quick Math Jr was hiding learning progress from players (it’s available to parents via a parental reporting system).

We assumed that kids would be uninterested in their learning progress or, worse, even find it demotivating. We assumed that the in-app reward system would offer feedback and motivation enough for kids to continue playing the app.

It took us an embarrassingly long time to realise the error of these assumptions!

By design, the early levels of Quick Math Jr are too easy for most 5 year olds. The games get harder but, because we’d hidden the progress system, kids had no way to know this. Eventually, enough kids complained about playing the same game over and over again for us to wake up and realise something wasn’t quite working.

As it turns out, 5 year olds both understand and are intrinsically motivated by learning goals. We recently tested a simple change to show kids their learning progress as they played. We didn’t have much of an idea about what to expect, but the results were amazing! Kids were much more motivated to continue playing a game when they could see that they were moving through levels of learning content.

We shipped the new progress indicators in our latest update and, what do you know, we’re seeing a 40% jump in playtime. Just one more reminder of the importance of listening to your user!

In short, testing sessions are extremely useful for learning about your product, and what is and isn’t working. However, and perhaps more importantly, they can also be great way to learn about your users and their needs.

Article by Jane Abrams, Shiny Things, Educational Content Development.

Stay tuned for the final post in this series (for now…), coming soon!