Emergence of Tech-Econ; Teaching Machine Learning, Big-Data Skills to Econ Students - Nazanin Kharza
#TeachECONference2021 Asynchronous Session Abstract:
Academia and industry increasingly demand big-data skills such as working with satellite data, generating and scraping data, visualizing data, and applying machine learning (ML) methods to economic questions. This paper guides adjusting our curriculum and effectively teaching these skills to undergraduate and graduate students. One potential reason academia has not yet adjusted to this demand is that few models are on such courses in economics to draw upon. The main challenge in designing such courses is to focus on economic applications while teaching the required coding skills. I summarize the existing resources and their use in teaching tech-econ and discuss how we can overcome this challenge by properly integrating economic applications and coding skills and selecting the right material. I discuss methods to develop a team spirit, teach students with different backgrounds, and train teaching assistants. I have taught big-data and ML courses at the University of Toronto and the University of Illinois. Students worked with real-world datasets and produced research papers that showcased their economic intuition and coding skills in these courses. My experience requiring active work on a research paper motivates students to have an ownership approach to the assessments and is essential for their success in a tech-econ course. Students learn and benefit the most from such courses by working on their research projects. Student feedback I received has repeatedly included comments like “I got the job because of the course project” or “the course project was a hit on the market.” The intrinsic motivation created by the substantial demand for economists with data science skills and the potential for meaningful research output makes big data and ML courses exciting to teach.