Jean Clipperton (Associate Director of MACSS and Associate Senior Instructional Professor), is a political scientist and computational social scientist who studies how institutions and individuals use language to construct shared understandings and identity. Her current research examines how emotional messages and symbols manifest and impact political messaging and campaigning, as well as how pop music encodes emotional meaning. She teaches courses in computational music analysis, data visualization, agent-based modeling, programming in R and Python, and research design.
Tell us a bit about the course context.
Computing for the Social Sciences is an applied introduction to data analysis and visualization with R that we offer synchronously online in the summer. Some students prefer it in person, but others appreciate the online option’s flexibility because they can pursue other opportunities, like internships or jobs. Students’ backgrounds vary, but most are undergraduates. Some are studying computational social science, others are fulfilling their quantitative requirement, and some are participating in the Summer Institute for Social Research Methods. Sometimes City College students take the course on exchange, and occasionally an intrepid Lab high school student joins us.
Tell us about how you engage students online.
Teaching on Zoom affords more opportunities for engagement than teaching in person. Since students are learning to use R in this course, we do a lot of in-class practice in which we work through pieces of code and coding problems together. I pose questions that they each answer in the chat, publicly or privately – for example, “what is this code’s output?” – and see a waterfall of responses. You can see who’s with you and who’s not. It can be challenging to get individual responses like that in an in-person class, but you can on Zoom. I also don’t require students to keep their cameras on, except for one row (in my Zoom display) of students, so that I can see some faces to gauge confusion, engagement, and boredom.
To examine students’ thought processes, I use breakout rooms and Google tools (Docs, Sheets, Slides, etc.) for in-class practice. I can see what everyone is doing and thinking as students complete practice problems together, working through each scaffolded step. If they need help, they note that in their document and highlight it so that I or the TA can assist. This approach lets me see how the groups are all thinking and lets me monitor more than in an in-person class because I can see all their work at one time.
How do you establish your presence as an instructor?
It can feel more difficult to get to know people online, but I want my students to get to know each other, and I want to form a community with them. I love getting to know students and their interests through office hours and in class. Sometimes I use warm-up Zoom chat prompts as icebreakers – such as, “what’s the best jelly flavor for a peanut butter and jelly sandwich?” People have pretty strong opinions about small things that they’re willing to share. I’m also not uncomfortable with silence; I don’t answer questions for students unless I know I didn’t ask the question well to begin with. Sometimes if I encounter silence, I tell students that the questions get harder from here. That usually encourages someone to speak up. I also highlight students’ voices during class conversations, without naming them if they submitted something privately.
What are the challenges in engaging students in online courses, and how do you respond to those challenges?
In-class participation and developing students’ investment in their learning are common challenges in online teaching. Early on I establish a status quo of high engagement by asking low-cost, low-risk questions in the chat – some are meant to help us get to know one another (“cereal is a soup, yes or no”) and others are low-stakes assessments using Zoom’s polls function. If students see other students participating, they’re more likely to do it. Additionally, attendance isn’t required. If multiple students are missing class, that tells me something about how they perceive class, so I try to make class time useful. Most students attend class even when meetings are recorded, so that indicates to me that students value synchronous activities.
What did you do to prepare to facilitate learner engagement in your class?
During the pandemic, while teaching at Northwestern, I attended several workshops with their ATS about Zoom and its many features. Even before the pandemic, active learning was central to my teaching because of its impact on student learning, so it remains central in my online classes. I also learned to use Google tools for collaborations, both online and in person with students, to conduct groupwork in class.
What is working well about what you’re doing? Why do you think it’s working?
The level of engagement is what I want, and it comes from the focus on student interests, letting students get what they want from the class via specifications grading, and emphasizing community. The class aims to show students how the content can help them pursue their interests, work with their prior knowledge, and fulfill their reasons for taking the class. The specifications grading approach lets them meet their own learning goals. In this approach, students pick the level of learning they want to achieve and work toward it, even if some assignments require a few attempts. Finally, I emphasize the community aspect of learning; that we all learn from each other, especially our mistakes.
What changes do you hope to make to your course in upcoming terms?
My evolving role in helping students think about their relationship with AI continues to impact the way I teach this course. I want to help learners avoid using it to short-change their learning. We might assume students always choose whether to use AI, but that’s not always the case; sometimes it’s accidental, like in suggested text and web searches. It’s everywhere and not using it requires intentional effort. There are also, of course, deliberate academic integrity violations. I don’t want to police students’ AI use; I just want them to understand its relationship to their learning.
Tell us how you think about assessment design in an online course, especially in light of AI use among students. What are the challenges, and how do you respond to them? How do you encourage students to think critically about their AI use?
As a game theorist, my approach to assessment relies on backward design – establishing learning objectives, designing assessments in which students demonstrate those objectives, and creating learning activities that lead students to accomplishing them. My course is about learning to code, and that pedagogical goal informs my AI policy. To help students understand how AI can both help and fail them, students complete an early assignment in which they enter bad code into AI, examine the output, and reflect on its helpfulness. Our AI course policy requires students to disclose resources, including ChatGPT, that contributed to their work. Novices aren't always able to evaluate AI output, so it’s important to show learners how to leverage it and what its limitations are. I can’t prevent them from using it, but I can show them how it’s unhelpful.
To learn more about specifications grading at UChicago, consider joining the Specifications Grading Exploratory Teaching Group (ETG), led by Borja Sotomayor, Senior Instructional Professor, Computer Science and Valerie Levan, Senior Instructional Professor, Humanities Collegiate Division. If you have additional questions about joining this ETG or about alternative grading in general, please contact teaching@uchicago.edu.