I'm co-chair of a task force at the University of Michigan to define computing education for undergraduates in our College of Literature, Sciences, and the Arts (LS&A) — see the start of our website here. We were charged by our Associate Dean Tim McKay, and my co-chair is Physicist Gus Evrard.
Michigan is an interesting place to explore computing education for liberal arts and sciences students. We already have a highly ranked Computer Science & Engineering division, which is my academic home. CSE prides itself on preparing "full-stack developers." We also have a School of Information, which offers an undergraduate degree popular with students interested in user experience design and in data science. So, if we have all that, what else does any student need — in particular, students who don’t see themselves as primarily focused on technology?
We’re currently in the information gathering phase of our process, and it’s been fascinating. The warmup activity for the nine faculty on the task force was to write a review of how computing education worked in their own departments. Even in LS&A departments that had strong computing themes in their practice, there were no explicit mechanisms in their programs to prepare students to use computing. The LS&A faculty knew where their students learned (for example) statistics, but weren’t really sure what computing was used (if any). We are searching course descriptions and program definition documents now to make a more definitive statement about computing education in LS&A.
We are developing a working set of definitions about what computing education means in LS&A. From our departmental reviews and interviews with LS&A faculty, the task force has identified three definitions that cover what LS&A faculty do with computing and what they want for their students:
- Computing for Discovery: The power of computing for discovery enables model building, data mining, computational statistics, numerical analysis, first-principles simulations, and machine learning/artificial intelligence applications to science, engineering and other disciplines. Increasingly, research discoveries are powered by advanced computing capabilities that allow manipulation and exploration of complex, often massive, data collections.
- Computing for Expression: Computing can be used to communicate and engage with others in ways that we couldn’t previously, which might include use and production of apps, virtual reality environments and games, social media, and the creation of new worlds that can be realized via simulation.
- Critical Computing: Computers are pervasive in our daily lives, and thus have immense cultural, social, and political influence. It is imperative to ask who is supported by computing, and who is oppressed by it, and how these outcomes impact the human experience — and whether we can create better models.
A theme that I heard in interviews with several LS&A humanities faculty was that they and their students struggle with understanding and using computing infrastructure. One professor told me that her students don’t understand how a website is different from paper from an archival perspective. A book once printed just continues to exist. She found that her students set up a website, and don’t realize the effort and costs involved in keeping that website available. She said that they don’t know what it means to pay for a server, or for a domain name, or what an ISP is. Another faculty who works with other digital humanities scholars says that she repeatedly has to explain about how to create accessible websites. When Gus and I presented our task force to the LS&A Student Government, the first reaction from several students who spoke was that LS&A students ought to learn how to use social media for marketing. When I related the story about accessibility, there was immediate interest in addressing the problem of inequitable access. Some of the students in the student government had never heard of screen readers or about the challenges of accessible Web design, and they saw this as an important topic that fit in many of their classes.
We're continuing our data gathering for another month or so, including a survey to all LS&A faculty — but then we get to the really hard parts. Meeting this variety of needs is not common in American higher education. Some of our peer institutions have a kind of quantitative or formal reasoning requirement where computing courses might appear as an option, but our exploration suggests that those courses only fit part of the Computing for Discovery definition. Computational modeling and simulation is about more than crunching numbers.
How do we provide the computing education opportunities that meet the needs of Letters, Sciences, and Arts students? And who’s going to teach these courses?
- Many sciences would like their students to know how to use computing for discovery. Should they each teach their own courses, or should there be some common ones? Are these statistics courses, computer science courses, or something new?
- Computing for expression is often about tools, but there are principles of design and fundamentals of how media are digitized that might be foundational for undergraduates. What computing foundations matter in computing for expression?
- Where should students learn about computing infrastructure, with an eye to critical consideration of whether it’s the way we want it in a just and equitable world?
- In how many of these definitions of computing is it useful to learn programming? There would likely be a need for different programming courses for different definitions. Are there advantages to teaching these all by faculty with expertise in computing education, or is it better to embed these within disciplines and programs and have them taught by the disciplinary experts?
These are challenges and questions that institutions in higher education will be asking over the next decades. Computing is the most flexible medium ever invented by humans. It’s going to be used in many ways in different disciplines. We’ve identified three themes of computing in LS&A at the University of Michigan. They overlap with the existing programs, but they have different goals. We have a lot to figure out to in order to meet the computing education needs of future liberal arts and sciences professionals.
Mark Guzdial is professor of electrical engineering and computer science in the College of Engineering, and professor of information in the School of Information, of the University of Michigan