Online teaching at higher educational institutions has become a much higher priority in the face of the COVID-19 pandemic, but most faculty and staff at these institutions are ill-prepared to adapt their teaching methods and content to this new medium. This article guides the reader through three teaching studios developed for online synchronous teaching to very different student populations: a large (90-student) graduate-level healthcare finance course at MIT, an even larger (200-student) undergraduate-level statistics course at the University of Tennessee, Knoxville, and a medium-sized (50-student) graduate-level operations management course at MIT. As we began building these studios, we found few applications in higher-education settings to rely on. Instead, we borrowed ideas and tools from the gaming community. Since different faculty will have different teaching styles and objectives, we have adopted a tour guide approach that describes the intent of each studio design, a complete listing of the software and hardware used in the studio, and a representative example of what the studio can achieve in practice. We conclude by documenting how other faculty have produced minimally sufficient studios for online teaching.
Keywords: online teaching, distance learning, livestreaming, massive open online course (MOOC), gamification
During the second week of March 2020, we were informed—along with most of our academic colleagues—that all of our classes were moving online immediately in response to the growing impact of COVID-19 in the United States. Although clearly the right decision at the time, the switch to online-only lectures was quite a shock for two of us. (Brian has been teaching online for years, so he was well prepared.) Despite having developed a considerable amount of online material in the form of massive open online courses (MOOCs), we (Andrew and Sean) had no experience delivering synchronous lectures or producing our own videos. In fact, we relied on the very capable services of various video production crews at MIT and the University of Tennessee, Knoxville.
So when the pandemic hit and we moved online, the three of us had very different experiences during the second half of the Spring 2020 semester. While the Zoom platform allowed us to reach our students remotely, the two of us with limited prior experience teaching via Zoom knew we weren’t reaching them emotionally. And thus began an intergenerational transfer of knowledge and wisdom about online teaching, from Brian to Sean, Sean to Andrew, and now the three of us to our colleagues with this document.
None of us are experts in audio-visual technology. Like most academics, our focus is research and teaching, not building home studios and livestreaming. However, we view our amateur status as a feature, not a bug. If we can do it, so can you.
Of course, Rome wasn’t built in a day, and neither were our studios. In fact, it took each of us a lot longer than we anticipated, precisely because we are not experts. Moreover, we’ve refined our setup over time, adding new components once we successfully implemented the basic elements of lighting, sound, video, and workflow. Also, much of the technology we’ve adopted is not made for professional videographers, but rather for video gamers, a decidedly more heterogenous population with less training and interest in audiovisual technology for its own sake. We figured that if they can do it, so can we, and that’s why you can too.
In fact, you likely only need a small subset of our studio components to support your specific teaching style and objectives, so treat this guide as a menu from which you can select your desired appetizers, main courses, and dessert—don’t order everything at once, otherwise, you’re likely to get indigestion.
Our goal is not to convert our colleagues to our way of teaching, but rather to write the tour guide that we wish we had when we started our journey into the “World of Edcraft.” Each of us has made a significant investment in online teaching technology, as well as a number of mistakes and miscues in applying that technology to our individual teaching contexts. We’d like to be able to share these experiences in the hope that it’ll save others some time and anguish, and improve online education for our students.
To that end, we begin in section 2 with an overview of our teaching philosophy and how it drove us to invest in our own home studios, and then describe in section 3 how online teaching can—at least in certain dimensions—dominate in-person teaching. We provide an overview of each of our individual studios and teaching formats in section 4, and then summarize in section 5 the most significant lessons learned from our experiences, including how to construct a ‘minimally sufficient’ home studio for under $1,500. We conclude in section 6 with a brief discussion of online teaching in a post-pandemic world. Appendix A documents the hardware we employed in our studios while Appendix B documents the software we used. A full studio list and illustration appears for Andrew in Appendix C, Brian in Appendix D, and Sean in Appendix E. And Appendix F contains resource links that proved particularly useful to us during this journey.
At the risk of being accused of academics, we begin with a brief and somewhat personal discussion about teaching philosophy. Not ‘Teaching Philosophy’—on which none of us is qualified to opine—but rather our own teaching philosophies, how we think about in-person versus online instruction, and why we’ve each made the decision to invest in considerably more complex technology than any of us would have thought sensible or possible a few years ago.
Also, the three of us teach somewhat different audiences, hence our teaching philosophies and styles reflect those differences. Andrew teaches primarily second-year MBAs who are broadly focused on management education, Brian teaches a wide cross-section of undergraduates, and Sean teaches first- and second-year MBAs with specialized interests in the more mathematically sophisticated fields of operations management and supply chain optimization. Each of these audiences has unique needs and expectations, and we hope this diverse range will span the particular applications of most readers.
In most other modern pursuits, the ability of technology to improve our lives has become a cliché. Moore’s Law, the observation that the transistor count on a microchip doubles approximately every two years (Moore, 1965), has transformed every industry, to the point where we now simply expect that each year will bring more powerful tools than the year before, whether it’s better smartphones, better cars, better medicines, or better video games. But what about better teaching?
If technology has allowed us to exceed our expectations in virtually every other aspect of our daily existence—computing, entertainment, retail sales, telecommunications, transportation, finance and insurance, health care, and so on—why hasn’t it had as much of an impact on education?
We believe the answer lies in the evolutionary history of Homo sapiens and our cognitive faculties. It’s a well-established fact that humans, and most other mammals, are social creatures. This truism has an important implication for our cognitive functions—we become much more alert and attentive in the presence of other humans. This feature clearly confers survival benefits: physical distance is highly correlated with the potential harm others can inflict. As a result, we become much more focused and engaged as another human gets closer to us. Phrases like ‘you’re violating my personal space,’ ‘you’re crowding me,’ and ‘too close for comfort’ underscore this basic principle of human biology.
This is why in-person meetings will always have an advantage over online interactions. Two-dimensional simulacra can only go so far in replicating a dynamic three-dimensional experience. And therein lies the challenge for online education.
The moment an instructor walks into a classroom and stands at the lectern in front of seated students, they command some part of everyone’s attention. If we measured the physiological characteristics of each student in real time—heart rate, breathing rate, blood pressure, skin conductance, and so on—we would observe a noticeable difference in these metrics as the professor arrives, begins to speak, and paces across the classroom and up and down the aisles. As the instructor approaches one side of the room, these metrics rise for students on that side and fall for those farther away. A skilled educator instinctively knows this and uses it to their advantage by approaching students to make a point, retreating back to the lectern to conclude that point and prepare for the next topic, and so on. Watching the pedagogical ballet of a talented teacher managing a lively class discussion is no less exciting than watching acrobats performing live on stage. OK, maybe it’s a bit less exciting, but you get the idea.
The point is that, if learning requires focus and attention, then in-person instruction will always dominate online instruction for purely biological reasons, other things equal. So the challenge of online instruction is making sure other things aren’t equal. To achieve comparable impact with online instruction, we need to use technology to alter the student’s perception of reality, in much the same way that the Hollywood film industry, the video game industry, the advertising industry, and the music industry alter our perceptions of reality with their wares.
A masterful exposition of the underlying scientific principles of how we learn is the book Grasp (2020) by our MIT colleague Sanjay Sarma, a renowned professor of mechanical engineering and successful entrepreneur who also happens to be MIT’s Head of Open Learning. Although the book was written prior to the pandemic, it reads like a survival guide for the past two years. There are too many insights to properly summarize here, but the following are key points that are directly relevant to the theme of this article:
Online learning offers pluses and minuses relative to in-person instruction, depending on the context, subject matter, and teaching approach. In a post-pandemic world, the goal is to take the best of both formats to produce an even more effective educational experience.
The ‘flipped classroom’ approach—where dense material is covered online via videos that students can pause and rewind as needed, and interactive lectures focus on applications and problem-solving skills—clearly works and this format should be more widely adopted and extended. Also, chunking video content into 10 or 20 minutes each can greatly enhance learning and retention.
Holding classroom discussion via Zoom offers the advantage of ‘equalizing’ all students so that no one gets to hide in the back row.
Recent research in the cognitive sciences confirms the importance of covering a given topic multiple times, and making the effort to integrate different parts of a lecture with each other. To maximize retention, students should be asked to return to a topic a week later, a month later, and then be given opportunities to relate that topic to others that they’ve covered earlier and later.
But perhaps the most important message of all is Sarma’s call for a new way of looking at education, not as a ‘winnowing’ process in which the best and the brightest are eventually separated out from the rest of the pack, but rather as a more inclusive, encouraging, and personalized process. By giving learners an opportunity to struggle with, and eventually succeed at, concepts in their own learning style and at their own pace, much better outcomes are more likely to be achieved for a larger number of students.
These ideas are completely consistent with the use of online teaching technologies described in this article, and provide a formal framework for motivating the use of these technologies even beyond pandemic lockdowns.
This challenge isn’t new. In fact, the video game industry has pretty well figured it out. Thanks in part to the pandemic, the global game industry’s revenue rose to $180.3 billion in 2021 (Palumbo, 2021). In contrast, global box-office revenues were $21.3 billion in 2021, but even prior to the pandemic, the highest-grossing year for the movie industry was 2019, with box-office revenues of $42.3 billion (Navarro, 2022).
Apart from the short-term challenges of filmmaking in the midst of a pandemic, why is the market for video games so much larger than movies? Both involve dynamic two-dimensional displays of images with sound and strong narratives that capture our hearts and minds. The key difference is engagement. While movies draw us in to the land of make believe, video games allow us to interact with its denizens. In fact, gaming companies like Activision Blizzard—maker of the Call of Duty and World of Warcraft franchises—have created entire ecosystems in which humans can assume various personas and engage in real-life drama with other human players. The popularity of these games has reached such heights that concerns about physical addiction have been raised by many sources. On January 18, 2022, Microsoft announced its intention to acquire Activision Blizzard for $68.7 billion, by far the largest deal in the history of the video game industry and a testament to the industry’s ability to engage consumers and generate recurring revenues.1
The success of video games in engaging players is no accident—video game designers spend significant resources in studying human behavior so as to optimize user experience (UX). One leading example is Dr. Celia Hodent, who holds a PhD in psychology and is a leader in applying cognitive science to video game design. In The Gamer's Brain: How Neuroscience and UX Can Impact Video Game Design (2018), Dr. Hodent begins with a summary of key neural substrates such as memory, attention, motivation, emotion, and learning, and then turns to various video game design specifications that are ideally adapted to these components. And she has taken theory into practice through her work at Ubisoft, LucasArts, and as director of UX at Epic Games. Online learning platforms have much to learn from video game experts such as Dr. Hodent.
In particular, there are at least five key functional aspects of successful video game design that are relevant for online synchronous learning platforms:
Measuring and motivating achievements
Allowing users to interact directly with creators
Capturing and maintaining user engagement
The old adage that ‘you can’t manage what you don’t measure’ applies just as much to video game achievements as to more conventional contexts, and anyone who has played a video game understands this fundamental principle. Therefore, gaming metrics such as the player’s number of ‘lives’ or level of ‘health,’ ‘energy,’ or other numerical score are a staple of every video game. Such metrics facilitate more specific achievements such as ‘top 10 high scorer’ league table status, transition to the next level, badges, trophies, extra play time or game privileges and powers, and other honorifics, all of which are designed to reward players and motivate them to continue playing.
Beyond achievements, players seek more direct recognition from key opinion leaders—often known as ‘creators’ and ‘influencers’ because they create original content associated with video games and exercise great influence on followers (e.g., livestreaming their gaming sessions, offering real-time commentary on other gamers’ sessions, providing hints and strategies for overcoming certain gaming challenges, etc.)—and the gaming industry facilitates such recognition through live and asynchronous video streaming platforms such as YouTube, Twitch, and Discord, as well as through social media platforms such as Facebook, Twitter, Instagram, and TikTok. A ‘shout-out’ or mention of a user by a creator confers significant pride and satisfaction on the user, who then develops a deeper connection with both the creator and the particular gaming platform or context. This is the postmodern instantiation of Andy Warhol’s dictum that everyone will be famous for 15 minutes, a commentary on the fleeting nature of fame as media evolved in the 1960s. Except in today’s videogamer environment of Discord, Instagram, TikTok, Twitch, and Twitter, every player will be famous for about three seconds.2
Engaging the player is, of course, the primary challenge of the video game designer, and their methods are well known (Hodent, 2018). However, in recent years, the rise of additional content that caters to spectators has created new revenue sources for the industry along with new technologies for engaging viewers as well as players. Dubbed ‘esports’ (short for ‘electronic sports’), this new aspect of videogaming has grown exponentially, not just through online participants but also with live events.3 Filmmakers have the same objective as livestreamers, but the latter have many more-powerful tools at their disposal to engage their audience, which is why the video game industry’s revenues have now eclipsed those of Hollywood. Immersing participants in a virtual world via challenging tasks that lead to rewards (in some cases, actual monetary rewards), providing recognition of achievements, punctuating tasks with breaks and scene switches to reduce the chances of monotony, and most importantly, maintaining constant two-way communication have proven to be highly effective at maintaining viewer engagement. We discuss this important topic in more depth in the next section.
Sustained player/viewer engagement leads to an even more valuable outcome from the gaming industry’s perspective, which is the emergence of community. By connecting multiple stakeholder groups through online platforms such as Discord, Twitch, and YouTube, the industry has facilitated a remarkably robust and active ecosystem, complete with its own stratifications of players, fans, and creators/influencers that have grown in popularity, diversity, and profitability that no one could have anticipated. Being part of a community provides stakeholders with a sense of belonging, family, mission, and ownership, binding them even more closely and durably to a particular video game.
Finally, the sheer complexity of today’s video games makes it difficult for any player or viewer to fully absorb all the game’s narratives and features, which creates the need for curation of content. This need has been amply fulfilled by creators and influencers who routinely take their followers on exciting tours of exotic video game landscapes and adventures that may only be accessible to elite players.
These five functional design elements can all be ported to online education. Achievements, badges, and league tables can easily be adapted to reflect academic performance. Students can connect directly with faculty via chat windows, shared screens, and social media, and they can be engaged more actively through polls, chats, breakout-room exercises, and other technology platforms. And community building occurs during livestreams via these real-time interactions—for example, through student moderators in the chat, allowing them to connect with each other and the teaching team—as well as outside the livestream through various discussion boards. Finally, faculty provide curated content both through their choice of lecture materials and via strategically timed breaks to delineate transitions to the next topic. We provide specific examples following of all of these applications of gaming technology to our online teaching experiences.
For all the positives gaming technology and design principles can add to online education, it is necessary to ensure that this format does not introduce bias in terms of who enrolls and how effective the course is for particular stakeholder groups. The problem lies not among the population of video game players, which was relatively balanced at 55% male/45% female in the United States in 2021,4 but rather in both the contents of video games and their designers and associated content creators and influencers. For example, Branson (2018, p. 218) observes that:
In addition to the content of the games, as an industry, gaming is more male-dominated than even information technology as a whole. Virtually no women occupy leadership positions. As to content, the games the industry produces are characterized by violence and cruelty, designed to appeal to adolescent males. The games feature women mainly as objects of sex or violence. The industry derisively terms the games and software it does produce for girls and young women as “pink ware.” It accounts for less than 5 percent of the industry’s output.
Although these issues may seem far removed from the application of video game design principles to educational contexts, there is a risk that incorporating gamification elements into the classroom can unfairly disadvantage some subpopulations or dissuade certain students from participating. For example, there is the risk that commonly used machismo-laced metaphors in gaming contexts may be carried through and inadvertently transmit gender biases to educational applications. These concerns led Brian—who, among the three of us, makes the most use of gamification techniques in his teaching—to introduce multiple ways for students in his online course to earn ‘points’ that would appeal to all genders such as game-show and trivia contests.
Addressing these challenges requires the vigilance and skills required to address these issues in a traditional campus setting, albeit reimagined for an online format. For example, anonymous polling allows immediate and unfiltered feedback from students, while activity reports document which students are not engaging with the course material. These tools help the instructor determine whether or not teaching objectives are being met, and if a course correction is required. Although none of the three of us is an expert on addressing bias in online learning contexts, we believe it is the responsibility of all faculty to be aware of this issue and deal with it as the occasions and opportunities arise.
Now that we’ve established that effective online teaching has a lot in common with effective gamer livestreaming, it should come as no surprise that much of the home studio equipment we’ve adopted is the same used by gamers. So how do gamers and video game makers maximize viewer engagement? From our admittedly amateur perspective, these are the key elements:
Continuous flow of action in sight and sound; never a dull moment, except for period breaks that provide punctuation for the action
Opportunities for two-way communication so viewers are also participants
High production quality
Narrative means a specific storyline or overarching theme to the livestream. Although gamers sometimes host livestreams where their viewers are just ‘hanging out’ with them, the more popular streams involve specific activities, for example, a competition, advice on how to defeat a specific boss, a tutorial on installing a piece of gaming equipment, and so on.
To maintain engagement, it’s critical to keep viewers occupied with a constant stream of input that fills as many of the senses as possible. Of course, even gamers need short breaks every now and then, so it’s important to establish a certain rhythm of action punctuated by short breaks, after which we return to the action, yielding a pattern of action, lull, action, lull, and so on that network-TV viewers are all too familiar with.
But the best way to keep viewers hooked is, of course, to allow them to join in the action. While Hollywood may have figured out ways to transport us to other worlds, video games allow us to live, breathe, fight, die, and regenerate in those worlds. In fact, watching a movie is no competition for being part of a movie and being able to influence its plot. This is what we need to do to teach online successfully, especially for business school audiences who expect greater interactions and engagement with their faculty and classmates.
We’re not the first to have come to these conclusions. In Harvard University’s hugely popular computer science course CS50, Prof. David J. Malan has managed to make one of the driest subjects come alive, not just for students in class, but for the many online learners who tune in synchronously. In fact, CS50 is the only Harvard course that students can watch live, in high definition, and Malan encourages students to do so, suggesting in a 2016 blogpost that it might be “a better educational experience to watch CS50’s lectures online than attend them in person” (Orbey, 2020). Malan’s example was one of the main inspirations for our online efforts.
In the remaining sections of this article, we describe the hardware, software, and studio setups we use in our online teaching. Now that we’ve stated our goals—which are much the same as those of gamers—the motivation for the various components of our studios will become clear. Each of us has also written separate case studies to accompany this article (Lo, 2022; Stevens, 2022; Willems, 2022) describing how we used our setups to suit our specific teaching styles and objectives in the context of the particular courses we taught online during the pandemic, so we encourage readers to refer to those articles for more details about our online teaching and what did and did not work.
We’ve also produced a series of videos about our studios as well as how we use them for online synchronous teaching:
Sean’s six-video series about his studio and teaching experience, https://bit.ly/3unclla.
We encourage readers who want to see, hear, and feel what it would be like to teach using our platforms to check them out.
One last point of clarification is in order. In describing the various technologies we’ve adopted, we decided not to write this document as a step-by-step user’s guide, but rather as a somewhat higher-level tour guide. The distinction may seem more a matter of semantics than substance, but it’s worth pointing out upfront because readers may wonder why we don’t provide more direction for how to set up a Stream Deck or a GoXLR device. The answer is simple: there are already many excellent tutorials on YouTube and other online repositories that do that,5 so why reinvent the wheel?
But a more important reason is that these technologies change so quickly that no matter what we write or how we write it, by the time you read this, it will very likely be out of date. For example, when Sean and Andrew first began their process of setting up their online studios in March (Sean) and August (Andrew) 2020, the main piece of software they used for broadcasting their livestream was Open Broadcaster Software (OBS) Studio, version 25.0.1. However, a much-improved version, 26.1.1, was released on January 6, 2021, and several key features have changed. One of them is that the new release no longer requires an additional plug-in to create a ‘virtual webcam’ that provides the video feed for Zoom. Had we written down the step-by-step procedure for how to use OBS to livestream into Zoom in August 2020, those instructions would be outdated in November 2021, when OBS Studio advanced to version 27.1.3.
The good news is that, given the size of the gaming community and how quickly it embraces new technologies, someone will have created a clear and often entertaining tutorial on how to use this new technology as soon as it’s available. So rather than write a user’s guide that will become obsolete as soon as we post it, our intention is to write a tour guide that gives you the lay of the land in the World of Edcraft, so you can locate the general neighborhood you’re seeking even if a few streets get renamed.
To illustrate why we believe that online teaching can dominate in-person experiences in certain dimensions, in this section each of us offers one example, describing the most useful teaching innovation we employed with our respective setups.
Because an important aspect of business-school education is interaction and engagement with one’s peers, the Fall 2020 online version of my Healthcare Finance course had to provide an unprecedented degree of real-time connections between students. This was accomplished via group exercises in Zoom breakout-room sessions. The exercises typically involved applications of a concept just covered in lecture, for example, how to compute the net present value of a drug candidate undergoing clinical trials, and took place in two parts. The first part involved sending students into preassigned breakout rooms of four to six students each for a 5-minute ‘icebreaker’ session near the start of class (Figure 1a), where they introduced themselves to each other. Then the students returned to class for additional lecture content, after which the breakout exercise was announced (Figure 1b), and then students were sent back into a 10-minute breakout session with the same preassigned teammates to work on the problem. At the end of the session, students rejoined the main class and I asked for volunteers to present their team’s solution to the rest of the class.
Figure 1. Breakout-room exercise slides for Andrew Lo’s Healthcare Finance course (15.482, Fall 2020). Slide (a) announces the 5-minute ice-breaker session and slide (b) provides the exercise itself, on which teams of four to six students collaborated in a 10-minute breakout-room session.
The purpose of separating the icebreaker from the exercise was to give students an opportunity to meet each other first and get introductions out of the way so as to be able to devote full attention to the exercise during the second breakout-room session.
The purpose for preassigning breakout-room membership was to maximize the number of new contacts each student made during the semester. In this respect, the online technology platform was able to accomplish something that in-person classes cannot: ensure that students are able to maximize the number of new contacts they will have made through this class by the end of the semester. Specifically, breakout-room assignments were made via an integer-programming algorithm that “samples without replacement” (Xu, 2020), so each week’s breakout-room assignment allowed students to meet as many new classmates as possible (Figure 2).6
Based on the final course evaluation for 15.482 administered by MIT (see, Lo, 2022, for details), this application of online technology was successful. Although students acknowledged that online interactions were inferior to in-person meetings, the lower cost of conducting meetings online provided much greater opportunities for networking, mentorship, experiential learning, and career development. To quantify the degree of networking, we tracked all class-facilitated first-time meetings throughout the semester between every pair of students. These first-time meetings are displayed in Figure 3, which is a ‘first-contact’ matrix where the -th element is colored green if students and met during a breakout-room session, practicum, moderator meeting, or informal meetings or office hours with me.7 Based on this data, the average number of new contacts made by students in the class was 43, the minimum was 35, and the maximum was 51. This suggests that a hybrid course—one with in-person lectures as well as some online components—could be superior to both in-person and online teaching.
Although I use a number of techniques to keep students engaged in Stat 201, the two components that I believe to be the most effective are “boss battles” and “speed runs” (see Figure 4). These features were borrowed from online gaming, and both work remarkably well at loosening up the class and keeping the lectures from being too static.
Figure 4. Screenshots of Brian Stevens’s “boss battle” (a) and “speed run” (b) segments from his online Stat 201 course at the University of Tennessee at Knoxville’s Haslam College of Business.
A boss battle involves my appearing as a menacing animated figure (see Figure 4a) using a similarly menacing distorted voice (courtesy of special effects by my GoXLR Audio Mixer, item B3 in my list of components in Appendix D), warning students that an upcoming topic is particularly challenging. It’s akin to video game challenges where a ‘boss’ appears and taunts the player, telling them ‘you’re going to lose, LOSER!,’ which typically has the opposite effect in motivating players to redouble their efforts to defeat the boss. For example, I use a boss battle when I’m about to cover Type I versus Type II errors, to get the students prepared by signaling that this is going to be tough, but don’t get discouraged—you can beat this boss!
A speed run serves a different purpose. Also borrowing from gamers who compete with each other to complete a given game as quickly as possible, a speed run is a timed teaching exercise in which I try to explain a concept as quickly as possible on a clean slate—usually a blank Word document that I bring up next to me. A countdown timer appears at the top of the screen to visually mark the remaining time I have left to cover the topic at hand. Most speed runs take around five minutes to present a concept that would normally take about half an hour. This might sound like a terrible idea from a pedagogical perspective, but in practice, it works surprisingly well. Repetition is the key to all forms of teaching, and speed runs are used as additional review after the material has already been presented on a prior occasion. Key concepts are explained in detail, but sample problems are not worked out, and the competitive aspect is whether I, the instructor, can provide a correct and articulate explanation of the topic within the allotted time. Students sometimes cheer at the end of the speed run and remark on whether a record was achieved. Not surprisingly, in each class there are a handful of skeptics hoping that I fail, which only adds to the drama and excitement of the event. With limited time to teach and the challenge of keeping students focused on material, speed runs provide an excellent tool for engaging online learners.
My primary teaching objective in the core operations management class is to give students the tools and confidence to solve operations problems they will encounter in the real world. Most students begin the course with the assumption that understanding the math is the challenge in the course. That is incorrect. The challenge is deciding the right problem to solve, and then developing the minimally sufficient solution that solves the problem. This requires rapidly presenting many simple building-block models to the students, so they can understand the breadth of models available and the relative strength and limitation of each model. We rely on the case method to teach the core operations management class, where a single case that highlights a specific model is presented every class. The course is already ‘flipped’ in the sense that the building-block models are presented outside of class. Class time focuses on the steps required to build the right model, develop alternatives, and communicate the recommended solution to decision makers. The course is 22 classes, and on campus it is common for students to feel quite uncomfortable with the format for the first 15 class sessions. In fact, when I solicit mid-semester feedback in class 11, several students will recommend that I slow down and take two class sessions per case instead of one. Even in class 11, students don’t realize they are progressing down the learning curve rapidly, and that, collectively, we will soon get to a point where they are comfortable focusing on how to use the models to solve problems, versus focusing on the math that underlies the models.
The primary ingredients for success on campus are getting students to ‘bog down in the details of the case’ and ‘immerse themselves in the moment.’ We need students to care about the details, otherwise they cannot directly link the models to the problem. We need students to immerse themselves in the moment, because this allows them to see that the model is a vehicle to quantify alternatives for the decision makers in real time.
My primary teaching innovation in the studio was to create an immersive experience comparable to the classroom. I switched the view the students saw through Zoom quite frequently, typically every 90 to 120 seconds. I moved between the model and student discussion rapidly so we did not overly focus on either the math of the model or the softer side of evaluating alternatives. And most importantly, we broke students into small groups where they worked together on exercises that extended the model in an unanticipated way. This let them see how dynamic these ‘simple’ models really are, and what they could achieve in practice.
Figure 5 demonstrates some of the views in one class session that calculates inventory requirements for an industrial manufacturer. After the initial model was articulated and built by several students, I presented an Excel spreadsheet that displayed the modeling framework with some of the key inputs hidden. We then thought through those inputs as a group before revealing them and the underlying calculations. Then we dug deeper into some of the math to better understand demand variability over different time scales; first in a PDF and then by hand to give students time to let the concepts sink in. More details can be found in Willems (2020).
Figure 5. Example of Sean Willems’s changing views. (a) Excel view with model inputs hidden; (b) split screen with document camera to define timing of events for model; (c) Excel view with model inputs revealed; (d) PDF that demonstrates mathematical theory behind model; and (e) split screen with document camera to write out modeling result.
Despite the differences in our approaches to online teaching, we all agree on one component that was so powerful, we felt compelled to include a discussion of it here, and that’s the chat function—the ability of students to send messages to either the entire class or individual classmates in real time, during class—which is available in one form or another on all online platforms (including Google Meet, Microsoft Teams, YouTube Live, and Zoom). This is perhaps the most novel aspect of learning technology from the perspective of traditional in-person teaching.
At first, we were taken aback by how much chatter was going on during our lectures. But we quickly realized that, while some of the chats were social—which helps build camaraderie among the students, a particularly valuable thing for an online-only class—much of it was questions and answers between the students. Because of heterogeneity in the students’ educational and career backgrounds, some students had deep expertise in the subject matter being covered, while others were totally inexperienced. The chat window allowed tremendous amounts of information to be communicated from one student to another and, in most cases, nearly instantaneously. Andrew provides an example in his video overview of his Healthcare Finance class in which he references ‘pharmacy benefit manager’ by its acronym, ‘PBM,’ without first defining it prompted one student to ask his classmates via chat what this term meant, and he received four replies—each with different and useful aspects of the definition—within the span of 22 seconds (see Figure 6).
On occasion, discussions in the chat become so relevant to the topic we were lecturing on, or, in rare cases, so confused, that the teaching assistant (TA)—who was asked to monitor the chat window throughout the lecture—would interrupt us to bring the issue to our attention so that we could engage the entire class in its discussion. This is a critical role of the TA, and their importance in maintaining an active and productive class dynamic cannot be overemphasized.
Once we realized the power of the chat window in crowd-sourcing information that usually supplements our lectures, we encouraged students to make use of this important feature. Moreover, during presentations by outside speakers, we would make use of the chat window ourselves, pointing out connections between speakers’ comments and course content, or posting articles to support points made by speakers or students.
For Brian’s online YouTube classes, chat is central because, unlike Zoom, the YouTube teaching platform does not allow video participants beyond the instructor. Therefore, the only way students can interact with Brian and their classmates is via the chat function. Therefore, Brian uses a dual-monitor display and the chat is always open on his confidence monitor, which is used to see what students are viewing live. Since the confidence monitor displays the livestream, the instructor can lecture from that monitor as well as moderate the chat discussion.
To keep chat active, Brian begins the class by greeting everyone via chat, and he uses a points system in Stat 201 that allows students to earn points by being active in the chat window (via an app known as MuBot). Students are encouraged to say hello or greet other students via chat while they wait for class to begin, and when the stream starts, Brian greets students individually as quickly and efficiently as possible. Many Stat 201 students enjoy this part of class, and consider it the 'shout-outs’ portion that builds camaraderie. In fact, some students have even asked for their kids to get shout outs at the start of class.
The basic approach that all three of us have taken in our online teaching efforts is to use software to create the particular format that suits our individual teaching style—'weatherman’ style for Andrew, ‘newscaster’ style for Brian, and ‘talking head’/lightboard style for Sean—and then stream this format using Zoom, YouTube Live, or some other videoconferencing platform (Figure 7). By separating the streaming component from the online teaching format, we’re able to produce richer (and more computer-CPU-demanding) presentations that students find more engaging.
The results of this approach are displayed in Figure 8, which contains screenshots of the different formats we use, and we provide brief summaries of each of our setups in this section. Andrew uses two primary scenes, one in which he stands in the middle of the frame when speaking without slides (Figure 8a), and a second in which he stands to the right of a slide using a ‘weatherman’ format (Figure 8b). Brian uses several scenes, but the common theme for most of them is a ‘newscaster’ format in which he’s sitting at a desk and speaking into a broadcast microphone with either slides (Figure 8c) or a homework problem (Figure 8d) as the background. And Sean uses the widest variety of scenes: standing in front of the camera (Figure 8e), standing to the left of a slide (Figure 8f), a split-screen scene in which he takes up the left half and a document camera fills the right half (Figure 8g), and a lightboard scene in which he writes on a transparent surface filmed with a black background (Figure 8h).
Figure 8. Three different online teaching formats used by Andrew Lo (a and b), Brian Stevens (c and d), and Sean Willems (e–h).
As we mentioned at the outset, the three of us have very different teaching needs based on the student audience and course content and, therefore, teaching styles. What works well for undergraduates would be totally ineffective for MBA students and vice versa. Even among MBA students, the ideal online teaching techniques for presenting mathematical derivations in a supply chain optimization lecture are different from those for presenting conceptual material on the ethical dimensions of drug pricing decisions. These differences gave rise to different designs for our individual setups and we believe this diversity will give readers a more complete set of options to choose from when designing their own studios.
Because Andrew’s preferred online teaching format is the ‘weatherman’ style, in which he is standing in front of his slides and able to point to specific areas during his lectures, his setup consists of an adjustable-height desk behind which he stands, facing his webcam, videoconferencing monitor, and studio lights (Figure 9a). The reason for his preference is simple: the weatherman format most closely approximates what a typical business student would see in an in-person lecture in a tiered classroom, making the experience less like an online lecture and more like ‘you’re there.’ In several experiments where he has switched between the ‘talking head’ and weatherman view, he’s noticed that the weatherman format seems to generate more class participation and back-and-forth discussion. He doesn’t know whether this is due to the influence of the format on the students or on him, but he’s convinced that this format provides the closest approximation to a typical in-person business-school lecture.
Behind him is a green screen, and off to the side are two smaller computer monitors mounted to swing arms and his desktop computer (Figure 9b). Sitting on his desk are the computer keyboard and mouse, a Stream Deck (a programmable set of keys for automating many functions such as quickly switching scenes in OBS, launching or closing programs, playing music, etc.), the GoXLR (for managing audio inputs/outputs and creating audio special effects like audience applause), the ATEM Mini (for managing video inputs/outputs), and the computer speaker on/off/volume control, all within easy reach. On the floor underneath the desk is the foot pedal control (for advancing/rewinding slides and other key presses by foot). Appendix C includes a list of all the components along with a photograph in which each component is labeled with the corresponding number from the list. Andrew used his home studio during the Fall 2020 semester to teach an online version of Healthcare Finance—13 three-hour lectures on Thursdays from 6:00 to 9:00 p.m. ET—and his experience with this course is described in more detail in his case study (Lo, 2022).
Figure 9. Online teaching home studios for: Andrew Lo (a and b); Brian Stevens (c); and Sean Willems (d–f).
Brian’s studio is a ‘newscaster’ setup (Figure 9c) in which he’s mostly in a sitting position—except for occasional trips to the whiteboard—and speaking directly into a broadcaster’s microphone, facing a tripod-mounted webcam and studio lights. On his desk are almost the same components as Andrew’s setup: keyboard and mouse, two computer monitors, ATEM Mini, GoXLR, and two Stream Decks. His desktop computer is tucked underneath the desk. Appendix D includes a list of all of Brian’s components along with a photograph in which each component is labeled with the corresponding number from the list. Brian has been using this setup for the past several years to teach online sections of Statistics 201 at the University of Tennessee, Knoxville, and his experiences are described in more detail in his case study (Stevens, 2022).
Sean’s studio (Figure 9d–f) is by far the most elaborate, consisting of a 420 square-foot room with two distinct camera locations from which he lectures: an adjustable-height desk similar to Andrew’s, and a separate area with a lightboard and its own webcam. At his desk area, he has a keyboard and mouse, computer monitor, a Stream Deck XL, and a document camera, and he faces a large videoconferencing monitor as well as a confidence monitor off to the side.
The lightboard station places the instructor behind a lightboard that is an illuminated glass board allowing students to see the instructor and the instructor’s writing at the same time. In contrast to the talking-head station, the instructor is always moving and visible at the lightboard station. There is an awe-inspiring aspect of the lightboard. It’s still novel enough that students have often not taken a class that uses the lightboard. Some subset of students is mystified by how the handwriting is not reversed. While one has to manage the space on the board properly, it is entirely possible to structure a board plan that is equivalent to the boards produced by a robust on-campus case discussion, as demonstrated in Figure 9f.
Appendix E includes a list of all of Sean’s components along with a photograph in which each component is labeled with the corresponding number from the list. He’s used this setup to teach Supply Chain Management to master’s students at the MIT Sloan School of Management throughout the Fall and Spring of 2020, and his experiences are described in his case study (Willems, 2022).
Online teaching has been a humbling, frustrating, and ultimately rewarding experience for us, and we wish we knew back in Spring 2020 what we know now, hence the motivation for us to write this article. In this section we summarize several of the most important lessons learned from that experience, including the six essential ingredients for an effective online studio (section 5.1), how to build a very good studio for under $1,500 (section 5.2), the importance of community (section 5.3), and a quantitative assessment of online vs. in-person teaching (section 5.4).
The differences in our three home studios reflect each instructor’s teaching style and preferences. As can be seen in our accompanying YouTube videos, each studio’s design decisions affect the way students interact with the instructor and each other. The reader will be well served to reflect on which studio elements embody what the reader will want to accomplish pedagogically. But as we reflect on the commonalities and differences between our approaches, we believe there are six core ingredients to create a successful studio:
A laptop or desktop computer
A dedicated microphone
A standalone camera that replaces the computer’s integrated webcam
A green screen or carefully chosen background
OBS to create the studio views that students will see.
A dedicated microphone is the single biggest improvement one can make to their studio. No other change will have as great a marginal impact as purchasing a dedicated microphone. The next three ingredients (dedicated camera, lighting, and choice of background) jointly define the teaching environment the students see. The interdependence between these three ingredients can make the purchasing decision overwhelming, but in reality, most any solution will work. As a means of analogy, purchasing these three ingredients is like purchasing a car. There are a bewildering number of choices that cannot be uniformly ranked across the buyer’s decision criteria, but most people settle into a brand or class of models and any of the choices will get the buyer from point A to point B. The final ingredient, OBS, is the open-source software that allows the instructor to create different scenes to reflect various teaching modalities the instructor wishes to use, including weatherman, talking head, and newscaster views. This ‘minimally sufficient’ home studio setup and the simple wiring diagram is illustrated in Figure 10.
Finally, as an upgrade purchase, we recommend an Elgato Stream Deck, which allows the instructor to switch scenes at the push of a button, thereby reducing lag time when transitioning from one teaching modality to another. This may seem like an unnecessary luxury, but we’ve found it to be invaluable in the context of synchronous teaching contexts in which students can easily lose focus and attention during ‘dead air time.’
Examples of studios that began with the six ingredients of section 5.1 are given in Figure 11, which are the home studios of three other faculty members in Business Analytics and Statistics at the University of Tennessee, Knoxville. Figures 11a and 11b are from Paolo Letizia; Figures 11c and 11d are from Justin Jia; and Figures 11e and 11f are from Chuck Noon. All three use laptops, not dedicated desktop computers. Any one of these alternatives allows faculty to provide their students with a much more compelling online experience than the standard talking head view, and at a much lower cost than the studios described in Section 4. Moreover, Prof. Noon put together his entire setup in 3 hours. Therefore, despite our own experiences, we believe it’s possible to improve your online teaching platform significantly with relatively little effort and expense.
Figure 11. Home studios of Profs. Paolo Letizia (a and b), Justin Jia (c and d), and Chuck Noon (e and f). All three studios use laptops rather than dedicated desktops, and cost less than $1,500 each (not including the laptop). Prof. Noon’s studio was assembled in three hours.
The primary elements missing from the studios documented in this section, relative to the studios described in section 4, are ease of use and permanence. The studios in section 4 have stationary, dedicated equipment and operations that are automated, and require much less setup time to produce the same experience every session. These features are the primary cost drivers of our setups. The same holds true for asynchronous teaching, that is, prerecorded lectures, which can cost $100,000 or more per course to produce professional-quality learning experiences such as those on edX, Coursera, and other MOOC platforms. However, very good-quality learning experiences can be produced at a fraction of that cost, as shown by UCLA evolutionary biologist Prof. Jay Phelan, an award-winning teacher who records and edits his own online lectures in his office, and the total cost of his setup is below $1,500 (Figure 12), as are each of the studios in Figure 11.8
Figure 12. Screenshots of YouTube instructional video for producing asynchronous lectures by UCLA Prof. Jay Phelan, an evolutionary biologist and award-winning teacher. (a) main scene of Prof. Phelan’s online lecture, with a list of his components; and (b) his setup for recording lectures. Source: Jay Phelan’s YouTube video, https://www.youtube.com/watch?v=znzECF9V-h4&t=835s (accessed 6 March 2021).
One common theme that all three of us learned in the course of our studio design and implementation is the importance of community in two different but closely related contexts: creating a sense of community with online students, and creating a similar community for online instructors.
To create community during a synchronous online session, having a dedicated TA is critical. In addition to dealing with administrative issues before and after class, which are opportunities to build rapport with the students, the TA plays a central role in moderating in-class discussions. For example, while we’re lecturing, we can’t easily monitor the chat window at all times, and on occasion some of the discussions in the chat become so relevant to the topic that the TA—who was asked to monitor the chat window throughout the lecture—would interrupt us so we could engage the entire class in that discussion. In this respect, the TA is the air traffic controller for each lecture. The obvious downside of online synchronous teaching from the TA’s perspective is the higher workload during lectures. A less obvious downside is that TAs remember virtually no content from the class itself because they’re simply too busy managing the participants to pay any attention to the content being covered.
Outside of lectures, TAs can also create community by providing timely responses to emails and holding Zoom meetings to address questions about lectures and problem sets. One of our most diligent TAs almost always responded to student emails within 30 minutes, so students seeking help wouldn’t be kept waiting and lose engagement. He set up dozens of individual Zoom meetings—some taking an hour or longer—outside of his regular office hours and weekly recitations to help students with problem sets and course projects, and quickly became an integral part of the students’ learning experience. Such responsiveness is, of course, highly dependent on the particular personality of the TA, but it would have been impossible without the online technology.
Developing community among online instructors is also critical. One rather mundane reason is that, every so often, a technical glitch occurs that will take us hours to sort out. And at the end of that process, we’re tired and emotionally spent, but triumphant in our newfound knowledge of how not to use the technology in ways it wasn’t designed to be used. If you enter this world of home studio development, be prepared to be frustrated beyond the breaking point from time to time. A certain Zen-like approach is probably necessary for preserving your long-run sanity. A more easily achievable solution is to develop a small network of like-minded colleagues—as the three of us have—with whom you can get and offer help and, more importantly, moral support. In short, you need a community of synchronous online instructors.
Our home studios may have a lot of moving parts, but none of them are particularly complicated. They’re all off-the-shelf technology that many people (mostly gamers, but a growing number of academics now) interact with every day. The two challenges to overcome are integrating them into a residential space and developing a teaching plan that best uses each component. Integrating them into the space is nontrivial because—assuming the instructor is not an audiovisual expert—small tweaks, like moving a camera 5 inches to the left, will affect desk placement, lighting position, sound levels, the place where the instructor should stand, and on and on. There are so many knock-on effects from changing any one component that integrating the technology components is time-consuming and tedious.
Even worse, many of these technologies don’t necessarily improve teaching quality; they only take the studio to the point where it’s just comparable to an on-campus classroom. Developing a teaching plan that optimizes what each modality of the studio offers is the more important challenge to overcome. It’s all about what the student experiences from the studio. The best way to employ this technology is a nontrivial task because it does not match how the instructor teaches on campus. It has all the requisite challenges of learning a new teaching format. Much like teaching on campus, there is tremendous value in learning from others and seeing what works. Both MIT and University of Tennessee at Knoxville hosted monthly workshops where faculty shared what worked for them, which was a tremendous opportunity to find a like-minded subset of instructors who were at the same point in their journey. Establishing a community through which ideas, challenges, and complaints can be shared is one of the best ways to encourage and improve online teaching.
Although we believe that online synchronous teaching can be as, or more, effective as in-person instruction, this belief is based largely on anecdotal evidence. However, in one case, we do have quantitative metrics with which to test this hypothesis: Brian Stevens’s Stat 201 course, which he has taught in several different formats from 2019 to 2021. He first taught a Zoom version in Spring 2019 (prior to the pandemic), followed by an in-person version during Summer 2019; then an online version using the YouTube platform in Fall 2019 (also pre-pandemic), both a hybrid version (due to the pandemic) and a YouTube version in Spring 2020, and a YouTube version in Summer 2021. There are, of course, differences in each rendition of Stat 201—if for no other reason than they take place at different points in time with different sets of students—but because the instructor, curriculum, grading policy, tests, problem sets, and virtually all other aspects of the course are identical, this is perhaps the best data we can offer to provide a meaningful quantitative comparison of online vs. in-person teaching.
The results of this comparison—which includes student performance and teaching evaluations—are summarized in Table 1 and Figure 13. Student performance is simply the scaled cumulative score of students for the semester, ranging from 0 to 100, which is used to assign their final grades. Teaching evaluations consist of student responses to 8 statements regarding the quality of their learning experience in the course, with evaluations on a scale of 1 (strongly disagree) to 5 (strongly agree).9
Figure 13. Graphical comparison of Stat 201 student performance and teaching evaluations across different teaching platforms—in-person, online via Zoom and YouTube, and hybrid—from 2019 to 2021. All sections were taught by Brian Stevensn at the University of Tennessee, Knoxville, with the identical curriculum.
The student performance results show that students performed significantly better in online and hybrid versions of Stat 201 than the in-person version. For example, the Spring 2020 hybrid version yielded a class average of 86.1, and the Spring 2020 YouTube version yielded 84.6, both of which are statistically significantly higher than the 74.2 for the Summer 2019 in-person version according to the Tukey-Kramer “honestly significant differences” (HSD) test.10 Given that the standard errors for these averages are 1 or 2 points, a 10-point difference in average performance between online and in-person courses is both statistically and pedagogically meaningful.
It should be noted that Brian is not the only instructor for Stat 201 in any given semester, and there are multiple in-person sections taught by other instructors using the same curriculum, exams, problem sets, and other course materials.11 This uniformity of course content across in-person and online teaching within each semester provides some assurance that Brian is not inadvertently adjusting his curriculum to be less challenging when he teaches online versus in-person.
Note. All sections were taught by Brian Stevens at the University of Tennessee, Knoxville, with the identical curriculum.
With respect to teaching evaluations, however, Table 1 tells a different story. Although the differences are not as pronounced as with student performance, the in-person course commands one of the highest ratings, 4.92, second only to the Summer 2021 YouTube version, which scored the highest evaluation, 4.97. The lowest rated rendition, at 4.67, was the Spring 2020 YouTube version; this was the first time Brian taught on the YouTube platform, so it’s no surprise that his second time was statistically significantly better according to the Tukey-Kramer HSD comparison. Our interpretation of these results is that in-person courses ‘feel’ better from the student perspective, but from the perspective of course performance, online instruction can be superior to in-person classes.
Although this is a small sample, and for only a single instructor, it does provide suggestive evidence that, at least in the case of Stat 201, online and hybrid versions seem to yield significantly better student performance than the in-person version, even though students seem to feel that in-person instruction is more effective. An open question is whether an intentionally hybrid course can dominate both in-person and online-only instruction (recall that the Spring 2020 hybrid version was due to COVID-19).
We recognize that a major investment in a home studio and redesigning a course to suit an online-only platform is not for everyone. In fact, the three of us have made such investments for different reasons, with different levels of knowledge, and at different points in our careers. It’s unlikely that junior faculty at research-focused universities would, or should, invest as heavily in online teaching given their other responsibilities and career objectives.
However, we’ve come to appreciate the incremental nature of progress in online teaching. We have added one component after another to our home studios, improving our platform after testing, tweaking, and occasionally cursing each addition as we struggle to learn how to use it. Our hope in writing this article and producing our respective YouTube videos is to allow our colleagues to pick and choose among the pieces of technology and teaching methods best suited for their individual teaching style. For some faculty, simply knowing about how the chat window in Zoom works and encouraging students to use it could dramatically improve their students’ class experience. For others, the six ingredients required to produce a step-change improvement in studio quality may inspire them to purchase a microphone. And for those faculty who share our passion for teaching and have the luxury of time and resources, a more substantial investment may be appropriate. In both cases, we hope this guide will lower their cost of entry into the World of Edcraft.
As of the date of this article’s completion, it seems that campus life is returning to a new normal and in-person lectures are resuming. However, given what academia has collectively experienced during nearly two years of lockdown and online instruction—and the economics of scale inherent in remote teaching—we believe there’s no going back. Although in-person classes are a welcome change for most of us, online versions will persist and may take on a life of their own through executive education, hybrid degree programs, and the next generation of MOOCs.
We thank Jay Phelan and Qingyang Xu for helpful comments and discussion, and Jayna Cummings, Amara Deis, and Rebecca McLeod for editorial assistance. Our thanks also to the HDSR reviewers and editors, especially Xiao-Li Meng and David Parkes, for their many helpful comments on the manuscript, and Allison Vanouse for preparing the video elements.
The views and opinions expressed in this article are those of the authors only, and do not necessarily represent the views and opinions of any institution or agency, any of their affiliates or employees, or any of the individuals acknowledged above.
Branson, D. M. (2018). The future of tech is female: How to achieve gender diversity. New York University Press.
Hayter, A. J. (1984). A proof of the conjecture that the Tukey-Kramer multiple comparisons procedure is conservative. Annals of Statistics, 12(1), 61–75. https://doi.org/10.1214/aos/1176346392
Hodent, C. (2018). The gamer's brain: How neuroscience and UX can impact video game design. Taylor and Francis Group.
Kramer, C. Y. (1956). Extensions of multiple range tests to group means with unequal numbers of replications. Biometrics, 12(3), 307–310. https://doi.org/10.2307/3001469
Lo, A. W. (2022). World of EdCraft: Teaching healthcare finance at MIT. Harvard Data Science Review, 4(2). https://doi.org/10.1162/99608f92.5c45c405
Moore, G., E. (1965). Cramming more components
onto integrated circuits. Electronics, 38(8). https://newsroom.intel.com/wp-content/uploads/sites/11/2018/05/moores-law-electronics.pdf
Navarro, J. G. (2022, March 16). Global box office revenue from 2004 to 2021, by region. Statista. https://www.statista.com/statistics/264429/global-box-office-revenue-by-region/#statisticContainer
Orbey, E. (2020, July 21), How Harvard’s star computer-science professor built a distance-learning empire. The New Yorker. https://www.newyorker.com/news/our-local-correspondents/how-harvards-star-computer-science-professor-built-a-distance-learning-empire
Palumbo, A. (2021, December 22). Gaming industry to mark another revenue record ($180.3B) this year. Wccftech. https://wccftech.com/gaming-industry-to-mark-another-revenue-record-180-3b-this-year/
Sarma, S. (2020). Grasp. Doubleday.
Stevens, B. (2022). World of EdCraft: Teaching statistics at the University of Tennessee Knoxville. Harvard Data Science Review, 4(2). https://doi.org/10.1162/99608f92.e303e4d3
Tukey, J. W. (1953). The problem of multiple comparison (Unpublished report). Princeton University, Princeton, NJ.
Willems, S. (2020). Designing a home studio for online synchronous teaching. https://seanwillems.com/wp-content/uploads/2020/09/Willems_DesigningHomeStudio_Final.pdf
Willems, S. P. (2022). Word of EdCraft: Teaching supply chain management at the MIT Sloan School of Management. Harvard Data Science Review, 4(2). https://doi.org/10.1162/99608f92.2b7d0bc7
Xu, Q. (2020, August 28). 15.482 breakout room assignment. Retrieved November 24, 2021, from https://drive.google.com/file/d/1bO7rUGJ7gXDZZMQiPtkgTGZjlAvSS0L2/view?usp=sharing
Let’s start with the hardware, which we divide into five categories: audio, video, computing, lighting and background, and miscellaneous. To provide some intuition as to how these parts are connected, we provide a wiring diagram in Figure A1 for Andrew’s studio.
These items are listed in Figure C1 (Andrew’s studio), Figure D1 (Brian’s studio), and Figure E1 (Sean’s studio). The components in each of the photos are labeled with numbers from equipment lists, and we use “(A5)” for component 5 on Andrew’s list, “(B7)” for component 7 on Brian’s list, and “(S13)” for the 13th component on Sean’s list. You can download the spreadsheet “components.xlsx” from our shared Google drive (see Appendix F), which contains links to every one of these components from the vendors we used (though we didn’t do a lot of comparison shopping so we have no idea whether we received the best prices).
Please note that these are just the components that we decided to use, and while they work well for our purposes, they may not serve yours. So, caveat emptor! You may want to pick and choose from this list and build your studio more gradually, incorporating one component at a time so that you develop sufficient experience before moving on to the next component. Some of them are quite expensive and represent a significant investment. Therefore, if you're looking to give just a few online lectures every now and then, it’s absolutely not worthwhile for you to do this. The only reason this kind of investment would make sense is if you’re going to be teaching online for a period of time and will be using this studio over and over again. We should also emphasize that we have no financial stake in any of these products or companies, nor do we receive any compensation of any sort for product endorsements.
For the microphone, Andrew and Sean use the Sennheiser wireless lavalier microphone (A13a). This is an expensive mic, but it’s phenomenal and we love it. It's wireless, which allows you to move around without any issues, and the sound quality is excellent. It uses the Sennheiser CHG 2 2-Bay Tabletop Charger (A13b) to charge the mic’s battery pack.
Brian uses the Shure SM7B Cardioid Dynamic Microphone (B1), and the sound is simply fantastic. Both Andrew and Sean have audio envy when they listen to Brian on Zoom calls because his sound is crystal clear, clean, and rich. Unless you spend a substantial amount of time listening to online content, you won’t fully appreciate just how remarkable this microphone is. Of course, the main downside is that you have to speak into it, so it won’t work if you want to move around. But for any kind of stationary lecture style where you don’t mind having a big microphone in front of you, this is the way to go.
The microphone needs to be connected to your computer, and there are two ways to do this. Sean uses the Sound Devices USBPre 2 Microphone Interface for Computer Audio (S6), which is a preamplifier that sits between the microphone and the computer. Its audio meters are very easy to read and make setting the gain totally idiot proof. Andrew and Brian use a different product, the TC-Helicon GoXLR Audio Mixer (A10, B3), which controls all the various audio components of your online lecture, including the wireless mic, your computer sounds, and creating special effects like the ‘Voice of God’ (changing your voice to be several octaves lower and with reverberation, which is particularly useful when making broad unsubstantiated claims). It also allows you to control the volume of different sources individually, so you can show your class a video clip that contains sounds, then narrate what the students are seeing by reducing the volume of the video while maintaining your mic volume.
The Sennheiser IE 500 Pro In-Ear Monitor (A20) is a set of in-ear monitors that allow you to hear exactly what your audience is hearing. This is particularly useful for avoiding ‘Zoom voice,’ the hoarseness that comes from straining your voice when talking to a computer or TV screen all day and unconsciously raising your voice to project the sound toward the screen. It took us a while to figure out why we were getting Zoom voice—which some of us were better at avoiding—but once we started using in-ear monitors, this behavioral tendency was greatly reduced.
We use several different webcams, most of which are digital single-lens reflex cameras (DSLRs) originally made for digital photography that have been repurposed as webcams, and they perform admirably. Andrew and Sean both use the Canon EOS RP DSLR camera (A6, S13), and Andrew has also used the Lumix G7 (a Panasonic brand and a favorite of YouTube vloggers because of its quality and price). Sean also uses the newer Canon C70 Cinema camera (S12), that has more features specifically designed for video. However, Brian chose to make the investment in a professional video camera, the Sony FDR-AX1 Digital 4K Video Camera Recorder (B5).
To turn the webcam output into a digital signal that your computer can understand, you’ll need some sort of conversion device. Andrew and Brian use the ATEM Mini HDMI livestream switcher (A5, B7). This is a very expensive component that does a lot more than just convert analog video signals into digital. For example, you can connect multiple cameras into this device and switch from ‘Camera 1’ to ‘Camera 2’ with a touch of a button. This is ideal if you want to broadcast two different views of the subject, or if you want a shot of yourself and one of your audience and to be able to toggle between the two instantaneously. However, for doing online lectures, this is probably overkill. You can easily get by with an Elgato Cam Link 4K device that Sean uses (S16)—currently $112.25 on Amazon—which resembles an oversized USB thumb drive. You connect your webcam to this device via the appropriate video cable, the device plugs into a USB port on your computer, and you’re good to go.
The Manfrotto 290 XTRA Aluminum 3-Section Tripod with 3-Way Head (A4a) is the tripod for the camera. On top of the tripod sits a remote-control pan/tilt head (A4b). With a solo operation, it’s very time-consuming to adjust your webcam’s position because you have to run back and forth between your desk and the camera multiple times before you find the perfect angle. With this device, you can move the camera up/down and left/right from your desk using the wired remote control, allowing to you point the camera exactly where you want it in a matter of seconds.
For showing documents and annotating them as you narrate, the IPEVO VZ-R document camera (A12) works well. For certain mathematical derivations, we’ll just put a blank piece of paper under the document camera and use a black magic marker to write the derivations as we narrate the specific terms we’re writing. Although you could easily present these same derivations in a nicely formatted slide, there’s something about handwriting mathematical expressions term by term that gives students an extra sense of comfort and understanding. Sean has a more elaborate setup—he uses the Canon EOS RP as his document camera (S13)—because he spends quite a bit of time writing content during lectures and wants higher resolution to better differentiate between writing instruments.
We each use different monitors, and the number and sizes depend on our individual setups. Brian has a smaller space for his home studio so his monitors are relatively close to him and are standard-sized computer monitors. Sean’s studio is much larger, hence his main monitor is the 82-inch Samsung RU9000 Series 4K UHD LED LCD TV (S17). Andrew’s monitor is placed about 20 feet from where he stands, so a Sony 65-inch TV (A2a) is sufficient.
We recommend purchasing two additional monitors (A16b), which may seem excessive but provide tremendous convenience and greatly reduce the mistakes you’re likely to make when switching sources, closing and opening applications, and sharing screens. One of these monitors will serve as the main screen for your computer, and the other monitor will be used to project your PowerPoint slides in presentation mode and for sharing your screen in Zoom.
Sean makes great use of the Height Adjustable Lightboard System, Extra Large (95-inch) (S21) produced by Revolution Lightboards. This is not an easy item to move, so getting it to his second-floor studio required hiring glass movers. Every detail of the design is thoughtfully considered, so from the instructor’s perspective this is plug-and-play, which is wonderful and a huge timesaver. The Lightboard Control Center (S22) takes care of all the hardware required to operate the lightboard with or without a computer, allowing the lightboard to look like any other camera source in Open Broadcaster Software (OBS) Studio.
All three of us learned the hard way that laptops just aren’t powerful enough for broadcasting an online lecture with all the functionality that we need, so we opted for desktops of different kinds. Andrew’s desktop computer is a ThinkMate VSX R5 540X1 workstation (with monitor, keyboard, mouse) (A16), and is the most expensive item on his list. Part of the reason is that it’s built not only for online lectures, but also for video editing and research computing (he conducts Monte Carlo simulations, maximum likelihood estimation, and machine learning on it). It has 256Gb of RAM, a 2Tb SSD boot drive, a 4Tb SSD drive for user content, and an NVIDIA Quadro RTX 6000 graphics card with 24Gb of memory, and is liquid-cooled so it’s totally silent (no fan noise whatsoever). Sean uses a Mac Pro with a 3.2 GHz 16-Core Intel Xeon W processor, 96Gb of RAM, an AMD Radeon Pro Vega II graphics card with 32Gb of memory, and a 1Tb SSD drive (S3). Brian uses an NZXT gaming computer that’s custom-built to comparable specifications (B13).
The Elgato Stream Deck XL (A1, B14, S16) is an indispensable device from the gaming community that allows you to assign multiple keystrokes to a single button, and the XL version of this device has four rows of eight buttons each, giving you 32 buttons that can be preprogrammed for all sorts of functions. In fact, you have many more buttons available because you can assign one button to bring you to an entirely new ‘page’ of 31 buttons (because you need one button to bring you back to the previous page), and you can have multiple pages of buttons. The Stream Deck’s software is integrated with several other streaming software platforms as well as OBS. This allows you to assign one button to the OBS scene with a full-sized image of yourself in the middle of the frame, and a second button to the OBS scene showing your PowerPoint slide in the entire frame and a smaller-sized image of yourself in the lower right-hand corner, ‘weatherman’ style. Switching from one scene to the next is literally at the touch of a button, and is instantaneous. This feature, more than any other, is what students comment on when acknowledging the ‘higher production quality’ of our online lectures, and is one of the most satisfying features of a home studio.12 You know this component must be really important if all three of us have it.
Even though you can hear sound from your computer via the in-ear monitors that we recommended above, you’ll also want speakers for those occasions when you’re not lecturing and you don’t wish to wear the in-ear monitors. We love the sound of good computer speakers (A15, S1). The Bose speakers (A15) have the added feature of a very convenient on/off switch and volume control that’s connected to one of the speakers via a two-foot cable, allowing you to place it next to your keyboard so you can control it easily while lecturing.
Finally, for those users who like to annotate their PowerPoint slides while they lecture, the Wacom Intuos Pro Graphics Tablet (A14) is the appropriate input device. We don't use this device very much because we find that the document camera provides a more dynamic platform for written input, but this is clearly a matter of taste.
Lighting is one of the most often neglected components in home studio setups, yet we believe it’s one of the most critical factors in creating a professional-looking online presence. Most video experts recommend using three light sources so as to be able to eliminate unwanted shadows and provide illumination for most setups. The NanLite Forza 60 LED Monolight (A3a) is a very compact but bright light and, when paired with the NanLite Forza 60 Softbox (A3a), provides soft diffuse light that doesn’t create shadows. The GVM 800D-RGB LED Studio 3-Video Light Kit (A3b) provides three very bright LED panels (of which you only need two) to provide more illumination for your set.
The ProMaster 10x20’ Solid Backdrop with the ProMaster Telescoping Background Stand Set (A7) is an excellent green screen that’s wide enough to accommodate virtually any shot you’re likely to need (this is especially important if you’re using the standard 16:9 aspect ratio for your broadcast, which is standard for HD video broadcasts). A green screen is essential for creating the ‘weatherman’ look, allowing you to replace your background with pretty much any image or video you choose. Although Zoom does have a ‘virtual background’ feature that also allows you to replace your background, it doesn’t work nearly as well without a green screen. Also, this Zoom feature adds an extra computational burden on Zoom, which can cause video lags and other glitches, so we’ve found that inserting a virtual background through OBS and a green screen is much more stable and visually convincing.
The final pieces of miscellaneous equipment include a desk from which to deliver your lectures, and then a few other important accessories.
For our desk, we wanted something that could accommodate both seated positions (for online meetings) and standing positions (for delivering online lectures), and was heavy-duty enough to hold the two monitors we recommended in section A1.2 as well as the workstation, the ATEM mini, GoXLR, and Stream Deck. Our recommendation is the Uplift V2 Commercial Desk (A11), an incredibly solid piece of furniture that allows the user to change the height at the touch of a button. This desk comes in many configurations and work surfaces, with many options including monitor arms, cable management systems, power strips, and so on. The only downside is its heft—the desk came in several large cartons and requires two people of reasonable strength to move into your studio space. From our personal experience, it’s possible for one person to assemble it, but once assembled, it’s still a good idea to have two people to maneuver it into its final position.
Any type of presentation will require a presentation remote, and we recommend the Logitech Spotlight (A17) instead of the more standard laser-pointer remote. The reason is simple: you can’t use a laser pointer for an online lecture, but the Logitech Spotlight works perfectly, creating a bright circle on a dimmed PowerPoint slide that moves as the remote is moved while the highlight button is pressed and held.
For occasions where you want to advance slides or press certain keystrokes without using your hands, a three-pedal foot switch (A9) is the solution. When would you want to do this? One example is if you’re prerecording a videoclip and wish to use PowerPoint slides like a teleprompter, in which case you need to advance the slides as you finish reading one slide and are ready for the next. Another example is if you’re interviewing a guest speaker and you’ve placed your questions and notes onto PowerPoint slides that you need to advance. A third example is the occasional need to locate your mouse pointer, which can require quite a bit of fumbling if you have three screens and are in the midst of a conference presentation or lecture. Just program one of the foot pedals to move your pointer to the middle of, say, screen 1, and you’ll never have to fumble again.13
This foot switch, and many other components we’ve listed, are connected to your computer via USB ports, but most computers won’t have enough ports to accommodate all of these devices, so we recommend adding more. The Sabrent 60W 10-Port USB 3.0 Smart Charging Ports (A8) is a good choice both because of the number of ports as well as the indicator lights and on/off switches for each port.
If you’re in an environment where power can be interrupted, or if there are regular voltage spikes—which are the enemy of audio/visual equipment, not to mention expensive computers—we highly recommend getting an uninterruptible power supply. Andrew chose the APC Smart-UPS 1500VA LCD 120V with SmartConnect (1000 Watts Output) (A18) for his studio.
You'll also need a number of cables and connectors of various types (A19), depending on your setup and the distances between components. If you decide to incorporate all of the components we listed into your home studio, you’re going to learn the differences between USB-A, USB-B, and USB-C connectors, and this is important because you’re going to need to buy additional cables and adapters to connect them all. Although each of your components will typically come with its own cable, you’ll run into situations where that cable isn’t long enough so you’ll need to buy extenders. However, you won't know what cables you’ll need until after you set up your system and decide where each component is best situated. Therefore, we recommend putting together your system first and making sure it works the way it was intended, and only after the setup is finished do you decide what kinds of extenders you need.
Online teaching involves a variety of software components which we group into the following categories:
Presentation content: PowerPoint, Keynote
Computation: Excel, Jupyter Notebook, R Studio
Video recording and livestreaming: OBS Studio
Video editing: Adobe Premiere Pro
Audio editing: Avidemux, Audacity
Video special effects: Adobe Premiere Pro
Miscellaneous Utilities: AutoHotkey, Keyboard Maestro, Snaz
We assume that readers are already familiar with presentation content and computation software, so we focus instead on the remaining software in this section.
In the same way that a Hollywood film studio brings together different elements to produce a movie—actors, writers, director, set designers, and so on—OBS Studio is a powerful piece of open-source software that brings together all the elements that make up an online lecture. Learning how to use OBS may seem daunting at first—at least, it was to us—but we found that after just a few YouTube tutorials, we were able to get the hang of it.
While OBS has many features and, therefore, many layers of complexity, the basic idea is straightforward: OBS takes various inputs—called ‘sources’—such as an audio feed (from your microphone, for example), a video feed (from your webcam or document camera), an image (such as a virtual background), or a display (like a computer monitor), and then combines them into a ‘scene.’ A scene is exactly what it sounds like using the filmmaking analogy: a particular arrangement of sources. For example, one scene might be you (the webcam) in front of a virtual background of a classroom. A second scene might be you in front of a PowerPoint slide show. And a third scene might be a split screen with you on the left side and the document camera on the right side. With each scene, you have total flexibility to control the position and size of each source, just as if you were a film director responsible for deciding how to shoot a car chase or love scene.
Once the scene is created, OBS can send it into a ‘virtual webcam’ that’s recognized by Zoom and once selected in Zoom, voila, your scene now appears to the other participants. The reason that more and more faculty are starting to use OBS is the tremendous range of control it gives us in creating different scenes quickly, easily, and with much higher quality than simply looking into a laptop webcam.
In fact, the ability of OBS to combine and process all these sources is one of the reasons we suggest using a dedicated desktop for online teaching. Most general-purpose laptops simply aren’t powerful enough to do all the things we might want to do in an online lecture.14 Of course, if you don’t plan to use more than a couple of sources and aren’t as concerned about occasional audio or video lags, then a laptop may work just fine, which is why we’ve recommended taking an incremental approach to building your studio over time.
After OBS, the second-most important piece of software is, of course, your online teaching platform, and we use Zoom (although Brian has recently switched over to YouTube for his live-streamed classes, so please check out his case study and videos to see how that platform compares). Because all the hard work in putting together the different components that make up a lecture have already been taken care of in OBS, there’s not a whole lot you need to do with Zoom regarding your streaming content (there is, of course, a lot to do with Zoom’s other features such as breakout rooms, the chat window, and so on, but we’ll get to that shortly).
Using OBS is especially useful for playing videoclips for your class. Rather than using Zoom’s screen-share feature to show the videoclip, you can bring the clip into OBS, position it exactly where you want (e.g., to take up the entire screen, or only part of it so you can show your students how you’re reacting to the video as they and you watch the clip, etc.). Because OBS sends the combined video stream into Zoom, there’s very little lag and far fewer skipped frames than if you used screen-share, even with the “optimize for video sharing” box checked. As long as the videoclip plays smoothly on your computer, OBS will provide an equally smooth stream into Zoom.
We also recommend using polls in Zoom so as to increase student interaction and participation. For any given 90-minute lecture, we’ll use two or three polls in which students are asked specific questions either about themselves (e.g., “I plan to pursue a career in healthcare after graduation: (1) yes; (2) no; (3) undecided”), or about the subject of the lecture (e.g., “The most I would pay for investment A is: (1) $1,000; (2) $2,000; (3) $3,000; (4) $4,000”). After the poll closes, Zoom allows you to show the results to the entire class, and then we spend a few minutes discussing those results.
We also use breakout rooms to allow students to interact with each other. As discussed in section 3.1, one very useful Zoom feature is the ability to preassign breakout rooms of any size, allowing you to group students according to specific objectives such as maximizing the number of first-time meetings between students in each breakout room (which Andrew used in his online Healthcare Finance course), or maximizing the diversity of students’ prior work experience within each room, and so on. The only drawback with this feature is that currently it’s only possible to have one preloaded set of breakout assignments per Zoom session.
Two other components of Zoom that deserve highlighting are the use of ‘raised hands’ for students who want to ask questions, and the use of the chat window so students can interact with each other. Both features facilitate student engagement with the instructor and, more importantly, with each other. In particular, the “wisdom of crowds” is a powerful source of information for any class with more than a handful of students, not to mention the camaraderie that can arise spontaneously if the right ‘chat culture’ is established at the outset. Because the instructor is typically unable to monitor the presence of raised hands or chats that should be brought into class discussion, the role of the teaching assistant during class is much more important in this context than in in-person lectures.
There is one issue with Zoom that instructors should be aware of, and it has to do with whether one should use Zoom’s screen-share function to show slides. The weatherman approach that Andrew favors, or the newscaster scene that Brian uses, eliminates the need to share your screen via Zoom. However, some—but not all—of our students reported a strange and frustrating phenomenon: when we don’t use screen sharing, the resolution of our image seems to decline, to the point where it becomes hard to read smaller-font text on our slides. With screen-shared slides, the resolution seems higher and the slides appear clearer.
This loss of resolution is no optical illusion. After a bit of investigation, we discovered that Zoom does seem to offer higher-resolution with screen-shared content than a typical Zoom session, and this policy seems to be institution-wide. Our assumption is that, to preserve bandwidth, Zoom keeps the resolution as low as possible, subject to some acceptable lower bound on the legibility of its stream. Given that screen-sharing often involves displaying text of varying sizes on a screen, the legibility constraint would call for higher resolution in these cases, hence the difference. This becomes most problematic for students with lower bandwidth connections, so whether your students complain about this issue depends to a significant degree on where they’re located and what kind of internet service they have.
Apart from asking Zoom to change the resolution on your account, the only two solutions we’ve come up with are: (1) to increase the font size used on your slides; or (2) to use screen-share mode rather than the weatherman format. We first encountered this issue in 2020, so by the time you read this article, Zoom may have addressed it.
So far, the software packages we’ve considered are meant primarily for delivering synchronous lectures. There are several other useful programs for preparing asynchronous lectures.
OBS does have a very useful recording function; in fact, that’s what all three of us used in preparing our YouTube videos of our online teaching setups.
After recording raw footage in OBS, we used several programs to edit that footage to produce the final versions of our videos. Adobe Premiere Pro, which is available for both Windows and Mac users, is our preferred choice for video editing, given its flexibility and ease of use. Mac users may also want to consider iMovie, which is used by many professional videographers, and there is also a Windows version.
For editing audio tracks, there are two excellent programs, both free and open source, that we’ve used extensively: Audacity and Avidemux. Avidemux allows users to edit both video and audio, and Andrew has used it extensively to fix a glitch in recordings made with his setup: a constant delay of 250 milliseconds between his audio and video tracks. Although this sounds small, it’s very noticeable and annoying when you’re watching a video with slightly mistimed audio. Avidemux provides a super-easy solution to this problem.
There are several helpful software packages that we use to support our synchronous online teaching, most of which are free. A countdown timer like Snaz is very flexible and is easily incorporated into OBS to create a screen that displays a digital timer which counts down the number of hours, minutes, and seconds to class time.15
AutoHotKey is a remarkably powerful free keyboard macro utility for creating shortcuts to accomplish a wide range of tasks. One example is a macro that moves the mouse pointer to the middle of a specific monitor. With multiple monitors, it can sometimes be a challenge to locate your pointer, especially if you’re in the midst of a lecture or Zoom meeting, so having a keyboard shortcut that can be assigned to either a foot pedal or a Stream Deck button is essential for being able to navigate your computer under stress conditions! AutoHotkey offers a scripting language for creating these shortcuts and the ability to compile these scripts into executable images that can then be run by many applications, including via the command line. For the Mac, there’s a more user-friendly proprietary utility called Keyboard Maestro that essentially allows you to ‘record’ a sequence of keystrokes and then assign that sequence to a hotkey.
To turn your smartphone into a webcam, OBS.ninja is a free web application that’s easy to use and works seamlessly with OBS so you can incorporate a second webcam source in your online lectures. This is what Andrew used to film his backup studio in his first video.16
Figure E1. Sean Willems’s home studio with labeled components: (a) talking head station; (b) lightboard station.
There are so many sources that we rely on regularly for designing, installing, configuring, and troubleshooting our studios that it’s hard to provide an exhaustive set of helpful links. We highly recommend consulting YouTube for any issue you run into, and chances are you’ll find exactly the right information within 15 minutes of diligent searching. The following are links to resources that we’ve found invaluable:
Andrew Lo’s home studio setup: https://youtu.be/jaFHhQktJiw
Andrew Lo’s video on teaching 15.482 online: https://youtu.be/hmGV_c-kriU
Sean Willems's home studio setup: https://bit.ly/2Lb7j9b
Brian Stevens's home studio setup: https://www.youtube.com/watch?v=IrP-0bAMlZI&feature=youtu.be and how he uses it to teach undergraduate stats: https://www.youtube.com/watch?v=P5N9_hvc9AI&feature=youtu.be
Hardware spreadsheet, and other files can be downloaded from: https://drive.google.com/drive/folders/1p31Za-YuU7AbGxiPjrsoN-oAc2uScihm?usp=sharing
To edit Andrew’s PowerPoint file, you'll need to install The Bold Font (used in the first slide only), which can be obtained at https://www.dafont.com/the-bold-font.font).
To create Andrew’s title slide with video behind cut-out letters: https://www.youtube.com/watch?v=PuSRHUehtCo
To download OBS Studio: https://obsproject.com/
Great OBS tutorial: https://www.youtube.com/watch?v=ySENWFIkL7c
Additional OBS advice: https://www.youtube.com/watch?v=CvzAoqXtREw
To set up the Canon EOS RP as a webcam: https://www.youtube.com/watch?v=y_2A0PYLNco&t=135s
To set up the GoXLR: https://www.youtube.com/watch?v=G2O_f5AGq7Q
To set up the ATEM Mini Pro: https://www.youtube.com/watch?v=AYjogcP-GXY
To create a countdown timer: https://www.youtube.com/watch?v=6NE3FDxJ_h0
To make your own stinger transition: https://www.youtube.com/watch?v=wGntVBSnA8Q
Patrick Winston's "How to Speak": https://www.youtube.com/watch?v=Unzc731iCUY
Jay Phelan's setup for producing lectures to be viewed asynchronously: https://www.youtube.com/watch?v=znzECF9V-h4&t=96s
GamingCareers YouTube channel (great resource for tutorials on all things audio/video/livestreaming): https://www.youtube.com/channel/UClx4eJ_EP9MJdz19JUjKD1w
EposVox YouTube channel (another great source of information for livestreaming tech): https://www.youtube.com/user/EposVox
©2022 Andrew W. Lo, Brian Stevens, and Sean P. Willems. This article is licensed under a Creative Commons Attribution (CC BY 4.0) International license, except where otherwise indicated with respect to particular material included in the article.