About Lana Yarosh

Svetlana “Lana” Yarosh is an Assistant Professor at University of Minnesota.

Peer Teaching Presentations in a Curriculum

I haven’t blogged last semester, as I’ve been devoting my energy to teaching my first class (which went great according to the student evals!). The class, titled “Design Methods for Computer Scientists,” was a project-based course covering inspiration methods (e.g., design probes), design process, sketching, and rapid prototyping (e.g., 3D printing, Arduinos). As always, I want to share my materials, so here is the syllabus and the project description. But in this blog post, I want to reflect on one specific component of the class, peer teaching.

The goal of this component was to help students practice independently learning new skills and teaching these skills to their peers. Each student was asked to prepare a 20-minute class session on a topic of their choice that may be useful to others in the class (see, peer teaching description and rubric). Most of the presentations focused on cool APIs, tools, and techniques for rapidly bringing design ideas to life. Topics covered included: Arduino eTextiles, littleBits, Intel IoT kits and Galileo, WebRTC, GreaseMonkey, Processing, PhoneGap, and many more interesting topics. The presentations were spread throughout the semester and graded both by me and by the other students in the class (40% of the grade coming from written peer critiques). In terms of the short-term goals, I was excited to see that the student presentations generally increased in quality as the semester went on, with them getting presentation tips from each others. Additionally, it was great to see that students incorporated techniques covered in peer teaching presentations into their projects, so they definitely learned through the process. The long-term goal of this component is to keep the class material from getting stale by continually updating the parts that are most likely to change quickly (e.g., tools and technology available for prototyping), but efficacy towards this long-term goal is still to be demonstrated.

At the end of the class, I asked student to rate the personal utility of each class component. Out of 13 class components, peer teaching came in right at about the middle of the pack as the 5th most useful. Through discussion with students, we considered how this can be improved and came up with some directions for the future. Next time I run the class, I’ll be asking for peer-teaching presentation to submitted as 20-minute tutorial videos and asking each student to watch 10 and critique videos that are most relevant to them. There are several advantages as I see to this approach:

  • The presentations were great, with the most helpful parts generally being the hands-on demos and walkthroughs. Even though presenters shared their slides, the demo portion of their presentations was ephemeral and students could not refer back to it.
  • Collecting peer teaching tutorials as video would help us build up a library of resources across years.
  • All the presentations can be due on the same day (making this more fair) and all of them can be available in time for the prototyping milestone. As is, about a quarter of the talks occurred after they could be usefully integrated into the students’ projects.
  • It would free up significant time in class, which would be best used for project workgroups. I underestimated the difficulty that student groups have in meeting up outside of class, adding in-class time that the all have to be there would really help. Next time I teach, I’d like to keep Mondays as project days.
  • Students who do a great job can share this tutorial widely and add it to their portfolio as an example of their work.

Has anybody tried something like this in their classroom? I would love to hear how that went! Stay tuned for more reflections from teaching, including some examples of cool projects that came out of the classes.

Better HCI Research in One Hour a Day: A Winter Break Guide for C.S. Students

Like many other schools, University of Minnesota has a fairly long winter break (more than a month!). The break from classes provides an excellent opportunity to catch up with friends and family, sleep, and (most importantly, of course) focus on research. If you are a Computer Science student starting or thinking about a research program in Human-Computer Interaction, consider if you can dedicate an hour a day to becoming a better researcher. If you’re interested, here’s a short (and I think, fun!) program to fill some gaps between an undergraduate Computer Science program and a graduate program in HCI.

Build a Psychology Foundation (10-25 minutes per day): Computer Science students who come into HCI frequently have the “Computer” part of HCI down, but have less exposure to formal studies of human behavior. One of the best ways to catch up is to take a crash course in Psychology. For this, I recommend the CrashCourse Psychology YouTube series. There are 39 episodes which (I think) get more relevant to HCI and Social Computing as the show goes on. Watching one or two episodes a day will give you a good start and set up a solid foundation if you want to explore any of the topics discussed in more detail. Here’s one of my favorite episodes from the show:

Start a Design Collection (10 minutes per day): Design plays an important role in HCI, but good design is difficult to teach. A big part of becoming a good designer is developing a practice of being open to inspiration, learning from examples of others, and sharing/critiquing ideas. One action you can take today to build this practice is to start a collection of inspirations. Whether this is a physical sketchbook or a digital collection in Evernote or Pinterest, the important thing is that you add to it regularly, share it and discuss it with others, and browse through it when you need ideas. If you have a specific research area already, start collecting images, articles, videos, and examples that are relevant to it. If you don’t have a specific area yet, collect examples of particularly good design and anything else that inspires you. As you build your collection, share it with friends and family — soon you’ll get people sending you ideas and inspirations and it’s a great way to reflect. If you want more information about design and collections in HCI, I suggest you take a look at Sketching User Experiences: The Workbook and the slides associated with the collections chapter.

Immerse Yourself in Good Research (25 minutes per day): Lastly, being a good designer or builder is not enough to be a good HCI researcher. One way of starting to develop good research intuitions is to read examples of strong research. The best paper awards at your target conference are an easy place to start (e.g., Best of CHI 2014). Pick one paper and spend 20 minutes reading it (it’s important to learn to read quickly, which may mean skimming some sections). As you read, note how the authors designed, framed, and described their work. Add the paper to your bibliography manager and jot down some of those insights for later inspiration.

By spending just one hour a day during this winter break, you’ll return to school in a much better position to do strong HCI research. Research is about building a practice. Consistent effort will look very much like genius after some time.


 

Late-Breaking Addition: If you have another 30 minutes a day, Shaun Kane suggests brushing up on your statistics by taking Jacob Wobbrock’s Practical Statistics for HCI self-guided course. It’s a 10-unit course, so doing two units a week will get you just about done with it by the end of the break.

Six Strategies for Including Children as Stakeholders

Adults are not good proxies for understanding the needs and experiences of children. And yet, I see the “proxy” approach surprisingly frequently in studies of home and family, health informatics, education and other HCI domains. I’ve recently written a case study on this topic for a new edition of an HCI textbook (Baxter, K., Courage, C. & Caine, K. 2015. Understanding Your Users: A Practical Guide to User Research. Methods, Tools and Techniques. Morgan Kaufmann) and I want to share some of the ideas here.

When seeking to understand families and children, it is critically important to include children in the research activity. While it may be tempting to use parents as proxies for gauging the family’s needs, this is insufficient for understanding the complexities of the family dynamic. In my own work, I’ve seen many examples of non-consensus and many instances where parents’ perceptions of their child’s experience diverged from the child’s. I provide several examples in the case study, but here I instead want to focus on six specific strategies to help researchers include children in user studies:

  • Working with children requires special considerations while preparing the protocol and assent documents. Children are not able to give informed consent, which will have to be obtained through their parents, but the assent procedure gives them the opportunity to understand their rights and what will happen in the study. It is important to emphasize to the child that they can withdraw from the study, decline to answer any question, or take a break at any time. In designing both the assent document and interview protocol, keep specific developmental milestones in mind. In my experience, I have been able to interview children as young as 6, however I always check the comprehension level of the protocol by piloting with friendly participants in the target age group.
If you interview in a lab and you wear a lab coat, you're gonna have a bad time talking to kids.

If you interview in a lab and you wear a lab coat, you’re gonna have a bad time talking to kids.

  • To encourage children to be open and honest, the researcher should actively work to equalize power between the child and the researcher. Children spend their lives in situations where adults expect the “right” answer from them. To encourage the child to share honest opinions and stories, the researcher needs to break through this power differential. There are a number of details to consider here: choose an interview setting where the child has power (e.g., playroom), dress like an older sibling rather than as a teacher, encourage use of first names, and let the child play with any technology that will be used (e.g., audio recorder) before starting. Above all, emphasize that you are asking these questions because you do not yet know the answers, that the child is “the best at being a kid,” and that there is no wrong way to answer any of the questions.
  • Parents make the decisions in a study that concerns their children, which introduces a unique constraint. As a researcher, you will have to respect the parents’ decisions: you may not be able to interview the child separately and you cannot promise the child that anything will be kept private from the parents. However, you can explain to the parents why it is important for the child to have a chance to state their perspective in private. In my experience, most parents are willing to provide that private space, especially when the study is being conducted in their home where they worry less about the child’s comfort.
  • Children may struggle with abstraction, so ask for stories about specific situations. For example, instead of asking, “how do you and your dad talk on the phone when he’s traveling?” ask, “what did you tell your dad last time he called you?” It may take more questions to get as all the aspects you want to discuss, but it is much easier for children to discuss things they recently did rather than provide an overall reflection. This is most important with younger children, but is a good place to start with any participant.
Children talk more during show-and-tell than just to answer questions.

Children talk more during show-and-tell than if you just ask them questions. (That’s my brother, by the way!)

  • Additional effort may be necessary to engage a shy child and one way to do so is to encourage the child to show-and-tell. For example, “Show me where you usually are when you think about your mom?” or “Show me some apps that you use with your dad on your phone?” Use the places and objects shared as stepping-stones to ask more nuanced questions about feelings, strategies, and preferences.
  • An example of a child's drawing of technology (a holograph robot for communication).

    An example of a child’s drawing of technology (a holograph robot for communication).

    Lastly, incorporating drawing and design activities may help the child get into the “open-ended” nature of the study, be willing to be a little silly, and reveal what may be most important to them. For example, I asked children, “What might future kids have to help them stay in touch with their parents?” These drawing are not meant to produce actionable designs, but will reveal important issues through their presentation. Listen for key words (e.g., “secret” = importance of privacy), look for underlying concepts (e.g., “trampoline” or “swimming pool” = importance of physical activity), and attend to common themes such as who would be interacting with their future device, where, and how often.

I hope that these six strategies demystify some of the processes of including children in a study. Please, let me know if you have any additional advice or any experiences in working with kids that you’re willing to share.