Tinkering towards semi-seamless hybrid classroom discussions
It’s not a good time for schools, at any level. Decades of chronic underfunding, even prior to the pandemic, have left them understaffed, overcommitted, and under-appreciated. The pandemic only exposed the public’s disregard for school funding, while placing greater strain in teachers to sacrifice not only livable salaries and time, but also their health, wellness, and even lives. It’s not sustainable for anyone and yet few in the public seem to have any appetite to adequately fund the institutions in which youth spend most of their waking hours, make most of their friends, and learn most of their knowledge and skills.
As schools have begun to partially reopen, the inevitable pivot hybrid learning was like a punch to the gut after an already brutal remote learning fight. Teaching in-person is already one of the hardest jobs in the world. (I say this as someone who has done it and also many other hard jobs, such being a Chief Technology Officer, working in a lumber yard, and recycling people’s nasty cans and glass bottles). Teaching remote, and in a pandemic no less, is even harder, as it stripped away so many of the critical social cues essential for orchestrating a group of students’ attention toward learning. But teaching hybrid is, by far, the most challenging: it combines all of the challenges of in-person teaching and all of the challenges of remote learning with no increase in a teacher’s attentional capacity and an added layer of audio/video configuration challenges. Asking any teacher, in primary school, middle school, high school, or college, to learn to do hybrid well while also worrying about getting infected with a once-in-a-century deadly virus, is simply too much.
Despite all of that, I believe that hybrid teaching is essential to our future. It is central to achieving equity, as not all students can attend school every day, for reasons of illness, injury, disability, caregiving responsibilities, or other factors that make commuting to a school temporarily or permanently infeasible. It is likely essential to public health, as COVID-19, influenza, colds, and other future viruses will always be a threat and one we’ll need to partly manage by temporarily isolating both contagious and at-risk students. And with the increasingly urgent threat of global warming to every facet of human life, the last thing we need is to further invest in the toxic carbon output of constant movement of billions of people back and forth to buildings. We have to figure out how to make hybrid work — maybe not right now, especially amidst a global pandemic—but eventually.
I have the privilege of experimenting. As a tenured professor at a research university, teaching is only a small part of my job, so I get to focus only generally one class a time, taught only a few hours a week. I have access to flexible classrooms. I’m compensated well enough that I can experiment with buying A/V equipment, even when the university is not yet investing in hybrid learning. Since I’m at a university that values exploration and academic freedom, I can freely try new learning and communication technologies. And because I am tenure track, how I spend much of my time is up to me.
And so I decided to spend much of it in the past two weeks trying to figure out how to teach hybrid well, building upon some of the nascent brainstorms I’ve written about previously.
The setting
Here’s my context. I’m teaching a graduate course to 35 students in on our Masters in Human-Computer Interaction and Design. The course is called User Interface Software and Technology, and covers what user interfaces are, how they are built, and the past, present, and future of user interfaces. I’ve taught the course four times before and so the curriculum and content is fairly stable, including an online book, with the same title of the course, that I authored to support it. Each day generally has the same structure:
- Starting with a question and answer session about a required reading to address student confusion and curiosity about the day’s topic …
- … then small group discussions about further reading about the topic…
- … then a whole class discussion about the small group discussions.
- … then an activity that concretely engages the concept of the day, and…
- … then a discussion about the activity.
And so the class generally involves a lot of talking: me sharing a point, a student posing a question, another student offering an insight or disagreement.
Pre-pandemic, the class was generally quite interactive. We would have fluid discussions while I walked around the room, with my attention on students’ raised hands, or their distractions on laptops or phones, or their non-verbal cues signaling curiosity, confusion, or intrigue. This rich social space worked pretty well, but relied heavily on my ability to constantly monitor 35 students’ engagement. I likely introduced the usual biases, overlooking students near the back of the room or not calling on students whose names I didn’t know. It probably reflected many K-12 classrooms in scale and complexity of classroom orchestration.
When I began planning hybrid for this quarter, the social space and constraints was quite different:
- About half of students were likely to be online, though this could unpredictably shift.
- Everyone in person, including me, would be wearing masks, removing critical social cues for facilitating discussion and muffling all of our speech.
- My classroom, a small design studio dedicated to these students, needed to have windows open to ensure proper ventilation, letting in significant road noise from the busy arterial street just outside.
This required very different strategies for both me and students. I could no longer use students’ faces for signals and my visibility into remote students would likely be more limited. Hands wouldn’t be easily visible to remote students as a cue to a desire to speak, a key way that students often use to regulate whether they participate. And all of the subtle participation cues like how many of the students’ gaze was on me or a student speaker, as opposed to a device, would be gone. How could I use technology to try to replicate some of these things, or create new kinds of information to facilitate instruction?
The setup
After much iteration, here’s what I landed.
The first realization I had was that I had two classrooms to design: the in-person one, which had to manage mask-muffled speech, road noise, and some window into the online classroom, to bring remote participants into the room. I a few key A/V choices resolved these issues.
A public address (PA) system. A brought a battery-powered PA system with a wireless mic and stood it next to the small table where I would stand. This amplified my speech, overpowering the road noise and dealing with the mask muffling. I could also hold the microphone when a student wanted to speak or hand it too them, amplifying their speech. Having just one microphone in the room, and requiring it to adequately hear anyone through masks, also simplified turn taking, as it basically guaranteed that only one person in the room was speaking at a time. Critically, the wireless mic’s range was only about 6 inches, so feedback issues were minimal, only picking up sounds in close proximity to the mic.
Projector and large screen. Behind me and the PA was a pre-mounted projector screen, bright enough to be visible in daylight. It projected the online classroom (which I’ll describe shortly), making any slides or instructions I wanted to share visible, as well as remote participants and other shared backchannels (which I’ll also describe shortly). I also encouraged students to have Ohyay open on their laptops, in case they wanted to have a more detailed, brightly lit view or contribute to chat.
Room-wide speakers. The projector was also plugged into room-wide speakers, allowing any of the audio coming from my computer — primarily the audio from the online classroom — to be amplified throughout the room. This was key, as any time a remote student was speaking in the online classroom, everyone could easily hear their voice booming throughout the room, just as the in-person participants’ voice would through the PA.
Laptop with front-facing camera and high quality ambient mic. I used my (quite new) 14" MacBook Pro, which has an excellent built-in ambient microphone array intended for conference calls. It’s also good at eliminating background noise. I stood in front of the laptop so that it would easily capture my voice from several feet away. Critically, however, I pointed the PA system speaker at the laptop, so that anything that was amplified through the PA would be picked up by the laptop microphone and routed to the online classroom. This meant that anything that anyone said through the PA mic would be clearly audible to online participants. Through the magic of the laptop’s background noise canceling, online participant speech routed to the room’s speakers did not feed back into the laptop mic, and so I did not need to worry about muting the laptop mic when someone online was speaking. And the laptop’s front-facing camera was always on, facing me standing in front of it, so the remote participants always had a view of my gestures (and whatever facial expressions were visible behind my mask).
An iPad with Center Stage. This device served as a video capture for the classroom, routing everything happening in the room to online participants. I chose an iPad mini because it was small but powerful and has great lenses. But a really nice feature of it is the front-facing wide-angle lens and Center Stage feature, which uses on-device facial recognition to find the faces in view and pan and zoom to ensure they are all optimally visible. It was like having an amateur cinematographer watching for who was speaking and automatically zooming in on them. And when no one was particularly in view, Center Stage showed a wide angle shot of the whole classroom.
A custom Ohyay classroom. Perhaps the most critical piece of all of this was my Ohyay room. Ohyay is basically an authoring tool for creating custom video chat environments. I’ve written about it before as a great online learning environment, particularly its use of emoji reactions. Some people use it for social gatherings, others for meetings. Because it’s such a flexible authoring tool, I’ve been able to use it to design highly-tailored online environments for the specific activities and pedagogy of each of my classes. Here’s what I designed for this particular class:
- A landing page, which explained to students the layout of the Ohyay space and also included some classroom previews, a place to ask anonymous questions, and links to key resources like our Canvas site for submitting assignments, our syllabus, and our asynchronous Slack for discussion outside of class.
- A classroom. This included several key elements: 1) prominent a video feed of my laptop’s front-facing camera, showing me, 2) a prominent video feed of my iPad mini’s front-facing wide lens camera, showing the classroom and whoever was speaking, 3) a grid of smaller video “slots” for remote participants, showing their video, name, optional pronouns, personalized emoji selection, and active speaking status, 4) three dedicated “speaking queue” slots, which students can move their video to if they want to speak, as an alternative to a virtual raise hand feature, 5) a screen share section, showing slides, web pages or whatever other shared content we were discussing, including student work, 6) a chat for both online and in-class participants to contribute to, creating a single shared backchannel for connecting around the discussion, or other social topics to facilitate bonding and culture, 7) and an emoji reaction bar, which served as an even lower barrier backchannel than chat, allowing remote participants to signal engagement, feedback, applause, and other emotions.
- Breakout rooms. We used these for small group reading discussions and activities. Each room had a limit of 4 participants, plus a 5th slot for the TA or myself if we needed to join and answer a question. It also included a preview of the main classroom so that students could refer to slides and monitor the chat. Each room also had a dedicated in-room chat.
There were several other small details into the room design that I think were subtly important:
- Playful fidgeting. I allowed students to freely rotate their video slots. When students were bored or feeling irreverent, they would spin their video slots around or turn themselves upside down. It gave me — and their classmates —a small glimpse into their engagement and their personalities.
- Room broadcasting. Whenever I needed to speak to a particular breakout room or all breakout rooms, as an admin, I could project my video and audio into all of their rooms and say things like “5 minutes left!”. As a nice touch, Ohyay also mixed together the audio feeds of all of the rooms I was broadcasting too, giving me and the in-class students a sense of the conversations and activity in the other rooms, just as they would have in class.
- Session recording. I set up automatic Ohyay recording on the classroom so that anyone who had to miss our synchronous session could watch our discussion. Students could also review it later if they wanted to review some interesting point someone had made.
I could manage most of this on my own, spending about 5 minutes setting everything up each day and 5 minutes tearing it down. But my TA also played a critical role online, making sure that the remote A/V experience was good, interrupting to highlight students in the speaking queue that I hadn’t noticed, and checking in on online breakouts while I checked in on in-person ones. She helped ensure the quality of the online experience, as I wasn’t able to easily observe it from a student perspective. (Although even without a TA, I think students could play this same role, if they felt sufficiently empowered to share feedback).
Reflections
I’ve only tried this for a week at this point and will likely keep iterating. But I think I feel confident making two claims from my experience:
- From a teaching perspective, it’s only incrementally more exhausting than teaching purely in-person or online. So much of the pain of hybrid is managing the complexities of A/V and attention, but the setup above managed to either automate much of that, or carefully design where my attention should be, so that the attention management wasn’t really more than a purely collocated class. Most of the time, I’ve felt myself naturally alternating between eye contact in the room, eye contact with my laptop camera, and monitoring the Ohyay speaking queue.
- Students in the class have said that it’s by far the best hybrid experience they’ve had. They’ve shared lots of helpful feedback about how to improve it, but many have said how surprised they are about how seamless it is. In-class students have shared how strongly they feel the presence of online students, and online students have shared how strongly the feel the presence of the classroom.
Of course, the setup isn’t perfect, it’s required a lot of new equipment, I’m building upon a lot of knowledge about A/V, user interface design, and pedagogy, I have a job that resources experimentation, and I happen to just be a fairly fearless person, willing to take pedagogical risks. So this is probably some version of a best case scenario for hybrid learning. Moreover, the setup above works for my particular class and its activities. It won’t translate to any arbitrary other class without modification. In fact, if there’s any takeaway from the above, its that designing any instruction always requires a mindful accounting of physical space, virtual space, social context, social cues, attention management, and more. Designing classrooms has always been hard—we’ve just long settled on certain patterns that make it a bit easier for teachers that don’t have the time or skills to design their own patterns, and we just don’t have those patterns yet for hybrid.
But we will. And experiments like the above are some of the necessary experiments for finding those patterns and then eventually building them into to our everyday classroom infrastructure. And so while achieving setups like the above today requires a lot of special expertise and risk taking, if we’re serious about equity, the setups of tomorrow will be built-in. Hopefully those in charge of designing classroom spaces take note and start resourcing more experiments, so we can find those patterns and build that future.
In the meantime, go tell a teacher you love them, that you respect them, and that you know their job is impossible right now. It’s never been harder to teach and so it’s never been more important to acknowledge the immense weight that every teacher is carrying right now, trying to balance their devotion to student learning with the safety of their families and selves. They need your grace right now, and in the future, your investment, so that when we resurface from this pandemic, we can build a more sustainable education system for everyone — including one that allows students to learn from home when they need to.
