My First Interaction, Design, and Children Conference, IDC 2018

I was visiting NTNU in Trondheim, Norway, last week, and attended IDC 2018, the Interaction, Design, and Children conference. Search Twitter for "#IDC2018" to get the tweet-by-tweet replay. This was my first IDC, though I’ve worked with colleagues who go regularly. I enjoyed it, and want to share some of my favorite moments from the conference.

The chairs were Michail Giannakos, Letizia Jaccheri, and Monica Divitini, and they did a great job. The keynotes (Pierre Dillenbourg, Kori Inkpen, and Alf Inge Wang) were insightful and inspiring. I particularly enjoyed Alf getting a whole roomful of academics to take a quiz on Kahoot!

IDC is a fun conference, filled with light-hearted moments—which is appropriate for a conference focused on interacting with children. Letizia invited a bunch of high school students to wander the posters and demos, dressed as fairy tale characters. There were lots of great videos and audio clips that had the audience break out laughing.

The best paper award went to a great paper by Lana Yarosh and team from the University of Minnesota, "Children Asking Questions: Speech interface Reformulations and Personification Preferences." I’ve known Lana for many years—she is a Georgia Tech alum, and we’d collaborated on a paper (on a terrific evaluation she did of our Media Computation data structures class) years ago. I was thrilled to be there when she got the award, and for the fascinating paper. They wanted to understand how kids used speech interfaces, and how changing the speech interface (like have it know their name) might influence how the kids used the interface. First cool part of the paper: They did it at the Minnesota State Fair! They had a booth at the fair, and found volunteers walking by. What a terrific way to get a cross-section of participants who might normally come by a University experiment. Second cool part of the paper were the findings and the careful analysis. They did the experiment with kids and adults, so that they could identify differences. The kids were very different. My favorites were, when the system said that it didn’t understand, when kids repeated the question louder or sang the question. Kids were more likely to include context and ask about that context, e.g., "Reggie was the fattest pig in 2010 at the MN State Fair. How many pounds was he?" And some of the kids thought that the speech interface that knew their name was "creepy." I wanted to know more about the kids’ and the adults’ computational model of the speech interface. They clearly had different models for how the interface worked. Some kids thought singing would improve the systems’ ability to understand them. How do they think Alexa and Siri work, such that they think singing would make thing better?

My favorite paper was "Designing for concreteness fading in primary computing" by Anthony Trory, Kate Howland, and Judith Good. Concreteness fading is a process for teaching abstract representations in stages (see here for a nice blog post on the topic). I know that Shriram Krishnamurthi and colleagues use concreteness fading to explain their work in Bootstrap. Trory and colleagues used some earlier experiments in mathematics to test concreteness fading in a CS context. They showed students a CS Unplugged-like activity about getting messages between islands (e.g., how long it takes to get from one island to another island), then broke it down into a more abstract representation about time between islands, and then finally described abstract routing tables with island names, times, and connections — fading from concreteness to abstractness. In their experiment, they asked students to route messages between islands in several conditions: Some students only saw abstract routine tables, some saw all concrete island maps, some saw reverse order from the theory (abstract, then intermediate, then concrete), and some kids saw the recommended fading concreteness order. It worked as predicted — students were able to solve transfer problems (about routing, given only routing table information) best if they got the concreteness fading condition. This is a pretty cool result for computing education, because it offers empirical evidence that the concreteness fading theory from mathematics education worked in a computing context.

The paper I’m most excited about was a poster "Game modding for computational thinking" by Marianthi Grizioti and Chronis Kynigos about their tool ChoiCo. They have kids play an on-screen game (the games looked like SimCity or a board game) that was authored in their authoring tool which is all built in Blockly, a blocks-based programming environment. They then let kids change the game Blockly. I’m starting a new project teaching economics with board games and simulation, and this is just the kind of thing I’ve been looking for! I’d like to see a layer where kids can define their own automated players, so that they can explore representing their strategy as code. Marianthi and Chronis are interested in this too, so I plan to be following this project.

The paper I learned the most from was "Overcoming socio-technical challenges for cross-cultural collaborative applications" by Sumita Sharma, Pekka Kallioniemi, Tomi Heimonen, Jaakko Hakulinen, Markku Turunen, and Tuuli Keskinen. They talked about the problems they had trying to evaluate their software in India which asked students to work with a guide in Finland to explore a Finnish city. It was hard to get the kids to be honest with the evaluators about what was wrong with the software — there was a sense of a "power distance," and the kids didn’t want to complain to the evaluators. (It reminded me of Gladwell’s podcastCarlos doesn’t remember.) So the evaluators invented challenges for the kids. The participants were willing to identify problems with the software when got in their way when they were solving the challenges. It was hard for the evaluators to understand the realities of these kids, especially the challenging lives of the female students who had so many expectations placed on them. A particularly poignant point was when the evaluator realized that she was asking the kids to help evaluate advanced technology, and these kids didn’t have socks in the winter. The cross-cultural boundaries made the research more challenging.

I don’t do a lot of my design work with children, but visiting IDC might convince me to do more so that I can hang out with this community: Great fun, fascinating discussions, and a lot for me to learn.