Premiere: Wednesday 19 June, starting at 3pm BST//10am EDT

---

Presenters:

How to Incorporate Small Activity-Based Lectures in Courses with Large Enrollments

John Hartman (University of California Santa Barbara)

With limited faculty resources, small class environments are difficult for many core courses. Although this is a major constraint, the use of technology can provide greater teaching efficiency. In this presentation, I provide a new approach to teaching introductory probability and statistics. With this approach, a single faculty member teaches and supervises graduate-student instructors using a flipped format, with most lecture content provided through asynchronous videos. This teaching format uses small-class in-person lecture activities, typically to 40-50 students per class, despite annual enrollment of about 1200. In order to make the workload manageable, additional support is provided by using auto-graded assignments, graduate Teaching Assistants and undergraduate tutors and graders. This approach allows students to discover that probability and statistics can be vibrant and answer important economic questions, such as addressing discrimination, whether students truly can randomize their choices when needed, and analyze student-collected data. While analyzing an assignment’s questions, students discover which topics are their strengths and weaknesses. Many students also realize during activities that the problem-solving approach is not clear in some instances. This results in much of the instructor-group interaction time in lecture devoted to addressing a process-oriented approach to problem solving. This approach lets students attempt to identify the key information in a problem, followed by understanding the question being asked. Once these steps are completed, students work with fellow group members (and the instructor as needed) to explore potential tools to solve the problem. In some cases, students realize the lesson that you sometimes have to fail before you find an approach that succeeds. In the end, students can self-evaluate to determine which parts of the process they do well at, while also knowing what parts they need to work on more.

Using Jupyter to Teach and Assess Econometrics

Jonathan Graves (University of British Columbia), Emrul Hasan (University of British Columbia), and Marina Adshade (University of British Columbia)

In this talk, we will present our work exploring the use of Jupyter to teach and assess learning in intermediate and introductory econometrics courses. Jupyter is an open-source, cloud-based system for creating interactive computational content (called “notebooks”) which is widely used in both industry and academia. It supports all standard computational languages, including R, STATA, Python, and Julia and has several advantages over more traditional tools. In particular, it facilitates the “literate computing” framework – which synthesizes analysis, discussion, and conceptualization to allow students to interact with econometric concepts and analyses in a way which is accessible to learners with different levels of expertise. We have used this tool in several sections at our institutions and are currently developing an open-source library of instructional tools for classroom use for our faculty members.

We will explain the tools we have used and how we integrate them into the classroom using some selected examples from different levels of instruction and content. The goal is to show instructors the possibilities these new tools offer, and how they can help improve student learning.