Dr. Gregory Wawro, PhD, Dr. Tian Zheng, PhD, and Dr. Andrew Gelman, PhD
Professor of Political Science
Professor of Statistics; Chair of the Department of Statistics
Higgins Professor of Statistics and Professor of Political Science
Dr. Gregory Wawro is a Professor of Political Science, Dr. Tian Zheng is a Professor of Statistics, and chair of the department of Statistics, and Dr. Andrew Gelman is Higgins Professor of Statistics and Professor of Political Science. In Fall 2020, they collaborated across different disciplines including Dr. Mark Hansen, David and Helen Gurley Brown Professor of Journalism and Innovation, to team teach an introduction to statistics course through the lens of the 2020 election. Given the remote teaching and learning context, the teaching team had to rethink how students could engage with the team and explore introductory statistics concepts through the lens of the 2020 election. Drs. Wawro, Zheng, and Gelman met the moment by leveraging their partnership and multiple modalities to support student learning. Read on to learn more about what they did in their course, what lessons and experiences they’re carrying forward, and the advice they have for other instructors at Columbia.
Establishing the Partnership
Dr. Wawro: This course was an introductory course in statistics for undergraduates, taught fully online. The learning goals for the course included developing skills for understanding statistics through the context of the elections.
Dr. Zheng: I taught this introductory statistics course for undergraduates, and referred to it as the “Election 2020 Edition” of Intro Stats. The idea of this course came to me as I was leading my department’s preparation for teaching virtually in Fall 2020. The course was STAT UN1101 Introduction to Statistics, the first course for our department’s Applied Statistics concentration/minor. The course aims to provide an introduction to statistical concepts, basic analysis, and visualizations. We had about 50 students enrolled in the class, many of whom were freshmen.
In Summer 2020, I was looking for ideas to make a Zoom classroom engaging and exciting for my students. Realizing that the 2020 election would be among the most consequential in U.S. history, I thought that studying the dynamics and outcomes of this (or any) election would be a great way to introduce statistical literacy. Excited by this idea and the value it would create for our students, a team was quickly formed: with me as the main statistical instructor, the course was divided into three sections, with each section led by a different faculty member with unique expertise: Dr. Andrew Gelman (Statistics and Political Science), Dr. Mark Hansen (Computational Journalism), and Dr. Gregory Wawro (Political Science).
Dr. Gelman: Statistics is a very general subject, with the same methods being taught, and used, in fields ranging from biology and medicine to economics and political science. When teaching statistics we always try to adapt the material to the interests of the students in the class. This time we set it up more formally by choosing the focus of application at the beginning so that students who were particularly interested in the election could select into the section.
Leverage Partnerships and Multiple Modalities to Enhance Student Learning
Dr. Wawro: To engage the students, we used a mixture of asynchronous and synchronous materials to cover a significant amount of ground. We separated the substantive part of the course into three modules, each taught by faculty with different areas of expertise (see section above) in order to keep the material fresh and engaging.
Dr. Zheng: Students were instructed to finish homework before each week’s classes, which were asynchronous learning materials on key statistical concepts. The class time was divided between “Election Talks” and “Data Labs. ”During the “Election Talks,” the students learned from our instructors and guest speakers about various aspects of past and current elections, many of which touched upon the use of quantitative evidence (i.e., data and statistics). During the “Data Labs,” students engaged in interactive, hands-on lab exercises using historical election data and data from election 2020, led by ten peer coaches in breakout rooms.
This course was well received by the students. They were engaged throughout the course, and the flipped format of the course and the role played by a managing statistical instructor and lab designer made it easier to engage researchers in such an introductory course.
Ultimately, I envision this class model as a template for other substantively-based and research-integrated introductory statistics courses. For example, one could easily see employing this model for a course based on public health, climate change, or misinformation.
Dr. Gelman: A colleague of mine once shared the following insight: Classes in the sciences are usually set up so that the goal of the student is to emulate the instructor and the authors of the textbook. But students can bring a lot into a course, with their own experiences, perspectives, and subject-matter knowledge. One advantage of teaching a course with four instructors is that it becomes clear to students that there is no single correct perspective on the material.
Advice for Instructors and the Future of Teaching at Columbia
Collaborate across departments and schools to deepen student learning.
Dr. Wawro: More collaboration across departments or schools can produce innovative and effective instruction, but there are both advantages and disadvantages to team-teaching with faculty members who have different areas of expertise; while it gives students multiple perspectives, maintaining pedagogical coherence is difficult.
Dr. Zheng: Statistical literacy is becoming increasingly important in a data-driven world. This kind of collaboration offers an innovative way to draw in students and keep them engaged in ways that are necessary to develop statistical sophistication.
Dr. Gelman: As faculty, we already collaborate across departments in our research and advising, and we already have students from all over the university taking our classes. So it is only natural for us to collaborate in the design and teaching of courses as well. We hope that, by spreading the word about collaborations such as this one, we will encourage other faculty to work together across disciplines to create new courses and revise old ones.
Leverage real-world experiences to engage students.
Dr. Wawro: It can be difficult to keep students engaged with remote teaching, but setting up online interaction helped to an extent. Whenever possible, create real-time exercises to illustrate points and keep students engaged; it’s good to use online surveys to keep students engaged and illustrate points. If teaching data analysis, use lots of visualizations but take your time in going over them with students.
Dr. Gelman: The best way of learning something is to teach it. So we structure the course so that, as much as possible, students can work in pairs. This allows them to explain ideas to each other and also keeps everyone engaged. It’s not so easy to stare off into space when you’re working with the person sitting next to you.