Getting Kids to Invent in a Giant Single-Day Workshop

Recently, I organized an invention workshop for AT&T’s “Take Our Kids to Work Day.” This involved three 45-minute workshops and almost 300 children (ages 7-15)! I’ve never designed with children on this scale and I wanted to share how it worked and some lessons from it (as well as share my materials, in case anybody would find that to be helpful).

What We Did: I gave a quick presentation on the 5 steps I take to invent, using the ShareTable as a concrete example. The kids were divided into 2 teams of 4 people at each table and each team had a different design challenge. They then had 20 minutes to come up with ideas and draw some inventions. Finally, they presented their best idea to the other team at the table, taking about 5 minutes each. I circulated throughout the rooms, focusing on the teams that were sitting back, instead of leaning forward.

What Worked:

  • Doing such large groups meant that I could very quickly get an understanding of whether something was working. For example, in the first workshop there were 4 teams that had the design challenge of a system that helps a shy kid who moves to a new school. All 4 teams really struggled with this challenge, so I was able to pull it out and replace it with different challenges for the two subsequent workshops.
  • Design challenges that focused on more physical ideas, like fun on car trips and taking care of pets, yielded a larger variety of ideas.
  • I got names and emails of families that might be interested in trying out new technologies. This is a great solution to age-old recruitment problem!
  • The prompt that worked best, directed to the whole group, was: “If you’re having trouble coming up with a good idea, write down a really bad idea. Cross it out and write the opposite of it.” See example below:

[Bobby] was thinking about ideas for better car trips. (1) He wrote down "I can't think of anything" in the middle of the page. (2) I make the suggestion that he think of a bad idea first, he writes "something that hijacks your car." (3) I tell his to cross it out and write the opposite of it, he writes "an app to help you find your car." He sees that as a good idea, gets excited and quick comes up with and draws 3 more ideas.

[Bobby] was thinking about ideas for better car trips. (1) He wrote down “I can’t think of anything” in the middle of the page. (2) I make the suggestion that he think of a bad idea first, he writes “something that hijacks your car.” (3) I tell his to cross it out and write the opposite of it, he writes “an app to help you find your car.” He sees that as a good idea, gets excited and quickly comes up with and draws 3 more ideas.

What Didn’t Work:

  • Very large groups meant very little one-on-one time with me. In general, the groups that I spoke to during the design session produced better ideas and were able to push past the initial “obvious” idea. Unfortunately, I couldn’t get to every team.
  • It was easy for the more shy kids to sit back and not participate, since there was little hands-on supervision from adults during the design exercise.
  • I was hoping to be able to do repeat sessions with kids who put down their names as being interested, however it seems that there are legal issues with having parents bring children to work outside of a formally organized event. So, it is likely that I will not be able to follow up with the children from this workshop in-person.

Why Do This: Designing with children is actually a great way to get ideas. Designing with a big group is great way to understand the ideas that would be valued by that culture. The kids are probably not going to come up with the next thing that you will patent, but with a little bit of translating, you can get to some interesting underlying nuggets. For example, take a look at these four ideas:

Four ideas from the workshop. These may not be directly implemented, but they can tell us a lot about designing in these domains.

Four ideas from the workshop. These may not be directly implemented, but they can tell us a lot about designing in these domains.

Now, let me try to translate what I got out of them: (1) We may not be able to make a translator for dogs, but perhaps there could be other ways of making the invisible visible, such as displaying physiological variables. This would be interesting as a contribution to the burgeoning field of pet-computer interaction. (2) Almost every team that worked on the translating problem came up with some variations on glasses and headphones (this one also came up with a typing glove), which shows that children might be quite comfortable with wearable computing, so that’s not a bad bet for the future. (3) Most teams that worked on the remote best friends idea, came up with something that was embodied, could interact with the remote space, and could participate in play. These are all excellent ideas to include in any technology for remote contact in the home. (4) There were a lot of variations on physical activities in cars (pools, slides, trampolines, etc.). While it’s probably impractical to go off and try to make this actually happen, this points to the idea that what kids really want to be able to do in the car is something physical. Hmmm… DDR for the car? *runs off to the patent office*

So to summarize, I think there is a lot of value to be had from even one-off invention workshops with children (as long as you’re willing to do some translating of the final ideas). Regardless of where you work, you may be able to organize one of these in the context of a take your kids to work day. Even if you end up with a giant group, it’s still doable and there are even some advantages to it.

Tips on Recruiting Study Participants

As HCI work expands outside of traditional computing fields and as we seek to design for users who are not like us, recruiting participants can become a really time-consuming and difficult process. I wanted to share a few approaches that have worked for me in the past, in particularly focusing on recruiting in situations where you can’t just sucker undergrads or lab-mates into participating.

  • Do your formative work (such as participatory observation or interviews) with an organized group, if possible. You may have to do some general volunteer work with the group first. This helps establish your legitimacy within this group and you will be able to continue working with both old and new members in the future.
  • Even if you don’t think that your potential participants have a relevant formal organization, try searching There are groups for just about everything and they’re usually happy to have somebody who is interested in their issue as a speaker, so it’s easy to make a connection.
  • Post widely on public sites. I have recruited a lot of participants through craigslist, which has the benefit of being fairly local. If your study can be done without meeting in person, I also recommend posting to forums that are relevant to your topic of interest.
  • Ask widely in your social network to see if anybody can recommend a participant for a specific study. Facebook is actually quite good for this task, but I’ve also found that bringing it up with people face-to-face gets people to think about it harder. I like to do a lot of looking on my own first, so that I can say “I’m having a hard time finding participants. Here’s what I’ve done so far. Do you have any other ideas?”
  • “Snowball Sampling” is when you get a participant to recommend other possible participants. I find that this works great! My one tip for making it even better is asking the snowball question twice: once when I follow up with the participant reminding about our scheduled meeting and again after the study is complete. This gives them a chance to think about it a little bit.
  • Do compensate your participants reasonably for their time and transportation. I have found that it is possible to recruit participants for free, but they often have ulterior motivations for participating which may clash with your study’s needs.
  • If your study can be done in one session and without special equipment (e.g., an interview), take advantage of the times when you are travelling. Just through posts or meetup groups or connections to friends, I usually get an additional 2 or 3 participants when I visit another state. For some reason, just because I’m there for a limited time, people feel more excited about being in the study (“You came all the way to CA to talk to me?”).
  • Lastly, if you need a small number of participants with very specific characteristics, I’ve found that it is worth the money to go through a professional recruiting firm. When I was in Atlanta, I used Schlesinger Associates and I was very happy with the results. I also think that in the end, it led to better data than using friends-of-friends, because the participants didn’t feel a social need to be positive towards my system. But, it is expensive.

One thing that I haven’t tried yet, but could potentially be interesting is using TaskRabbit, which is a site for posting quick tasks and having people in your area do them for money (so, it would only work if you’re compensating participants). If anybody has tried it, I would love to hear about your experience.

Here, Have Some Storyboard Templates

An example of a storyboard template sketch included in the file.

[Storyboard Templates File]

Just a short post this week. I’m settling into my new AWESOME job at AT&T Research Labs and figuring out what I should and shouldn’t blog (you know, IP and all that). In the meantime, I wanted to throw some storyboarding templates out, in case somebody will find them useful. These are ultra-sketchy, so you can customize with sketchy pics of your gadgets and interfaces. The file includes a PowerPoint with a 4-panel storyboard template and a few images that I’ve created and used in the past few months. These should be general enough to suit most needs, but if you really need something else, I can do a custom if you ping me a few days before you need it.

I found that these types of storyboards are useful in quickly communicating the basics of an interactive system to the user because they encourage presenting a compelling use case scenario rather than just wireframes of the interface. Since the boards are so low fidelity, you can get away with fairly vague specifications of the GUI. In my experience, this leads users to give feedback about how the system would fit (or not fit) into their lives rather than commenting on something like the color of the buttons.

These sketches are free for you to use as you see fit, but do let me know if you find them useful, as that is always encouraging to hear.

At Eugene’s request, here are thumbnails of all the template images you get:

Thumbnails of the 10 images in the template file.

[Storyboard Templates File]