An awkward class design (#1138)

Topics/tags: Teaching

This term, I’m teaching Human-Computer Interaction for the first time. For some complex reasons, we decided to make HCI [1] a two-credit course, which means that it should occupy about half the time of a four-credit course. I tried to keep the class reasonable; I cut readings and exams from the prior offering. Nonetheless, when I asked the students how the class was going by week three, they reported that it was taking more than half the time of a normal class and more than they were happy spending on the course [2].

That got me started thinking about things to cut. I know that setting up meetings with other people can be hard even in normal times, so I’ve tried to include times for project groups to meet. I haven’t always been successful.

But I’ve also been looking for other ways to cut the workload. The students’ current project [3] is to conduct a simple usability test on a Web site or app of their choosing. Rather than force them to round up research subjects and schedule time with those subjects, I decided to use class time for the interviews [4].

I figured it would work like this: During the first half of class, half the teams would run usability tests with the members of the other teams; during the second half of the class, they’d switch roles. Then I did the math. If they did fifteen-minute tests, we allowed three minutes between tests, and each team did three tests, we’d occupy essentially all of the time allocated to class [5]. That seemed fine. I waited until I knew who would be in class [6], and then I put together the schedule of interviews. I sent the schedule and instructions to students. They went something like this.

1. To fit everything within the time allotted for class, I am planning 15-minute interviews [8], starting right at 3pm, with a three-minute gap between interviews. Partners should plan in advance for their interviews, and plan to do the interviews as chats in Team. (E.g., one partner starts a chat with the other partner and the interview subject.)

2. Please plan to record at least one session so that I can view it later.

3. We will not meet as a whole class. I’ll send out the normal set of announcements after class.

4. [The class mentor] and I will be available to answer questions and help out during class.

Five minutes before class time, I posted a Get ready for class message in both the Class Meetings channel in our class Team and the class chat on Teams [9]. At the start of class, I posted a time to start your first interview message.

And then …

I realized that in an online world in which students are supposed to be working in small groups, it’s hard to tell what students are doing.

After fourteen minutes, I posted a one minute to go message. After one more minute, I posted an announcement that they should conclude their first interview. I added a few more notes.

But I saw no feedback. Were students actually meeting? I couldn’t tell. I couldn’t peer over their shoulders. I’d thought about asking them to invite me to their trials, but I realized that having me pop in and out might be distracting. I also worried that I’d need to do some behind-the-scenes management, such as responding to Our research subject isn’t responding, what do I do? messages.
So I found myself simply trying to keep track of time and send announcements—without knowing what was happening. It was a very strange experience.

As we started the three-minute break between the second and third rounds, I posted a reminder message as to what they should do next.

Get things set up for your third interview, which should start at 3:36 and end at 3:51.

I paused for a moment and, well, I posted what I’d call a Sammish comment [10].

(Sam wonders if there’s anyone out there.)

I got two likes for the Get things set up comment, as well as three thumbs up, three laughs, and two surprises for the second comment. I think there was only one overlap. I should have had twenty students in class. Should I take nine responses as positive or problematic?

In any case, I continued. I needed something to kill the time, so I started writing this musing. But I kept my eye on the clock, so that I could send out the announcements. I almost certainly made the right decision not to visit the studies. But I’m a busybody; I want to know how they are working [14]. I assume that the fifteen-minute time-frame feels tight. Running tests online is strange.

I hope some of them recorded their sessions so that I can still pretend to be a fly on the wall.

So I’ve now spent a two-hour class period running a timer, sending notes to students, answering the occasional question, and worrying the whole time. It feels strange. I wonder if I did the right thing for today’s class.

I’m finding myself wondering whether the next time I structure a class this way, I should find a way to observe. Or maybe I shouldn’t assign class time to these tasks. I’m not sure. It just feels strange.

Oh well. Teaching sometimes involves risks and trying new things. On Monday, I’ll debrief with my students. Maybe I’ll even muse about what they say.

Postscript: If I were a better teacher or a more sensible human being, I might have used class time to make a dent in my grading.

Postscript: It does feel somewhat right to experiment with this style in an HCI class. Maybe we’ll also discuss the design principles of today’s class design on Monday.

[1] Human-Computer Interaction.

[2] There may have been an exception or two.

[3] Maybe assigning weekly projects wasn’t the best idea. But we learn by doing.

[4] I suppose interview is not the right word. User study would probably be the sum total of all the individual studies. Perhaps session would have been better.

[5] 3:00–4:50, with a five-minute break somewhere in the middle.

[6] Sam is optimistic and believes that letters to the class that say I’m going to assign test subjects, let me know by 1 p.m. whether or not you’ll be in class [7] will be successful.

[7] Large portions of the class got their second Covid-19 shot the prior day.

[8] They knew what I meant. At least I hope they did. Otherwise, I’m going to get a lot of profiles of students.

[9] At some point, I discovered that it was easiest to get messages to my students who were scattered if I created a separate chat. I know that I could have set up a tag, but the chat seemed easier.

[10] I was at a meeting a few weeks ago. Someone asked Are you good, Sam? I gave one of my traditional responses, like Does it matter whether or not I’m good? I’ve completed my tasks. [11] They followed up with, Okay, are you Sammish? [12].

[11] I may have added, Traditionally, I think of myself as chaotic neutral.

[12] I believe we had a side chat about how one spells Sammish. We settled on the spelling I’m using.

[14] Or, as I worry, how they aren’t working.

Version 1.0 of 2021-04-30.