My academic program employs my dual perspectives as professor and executive to provide students with interdisciplinary learning opportunities in data science. The program’s motivation stems from my own professional journey. The students I interacted with as a hiring manager, including those who had earned MBAs and PhDs, tended to be hammers without a nail. It was hard to achieve a return on investment that exceeded the 20–30% premium I paid for their methods training without investing in additional after-hire training. I recognized in them the same challenge I had faced as an early-career software engineer: there was no bridge to data science. I had to forge my own. As student enrollment in data science programs has grown, I see an opportunity to show them the intellectual bridges my peers in data science and I had to build for ourselves—before data science was called data science.
When we fail to provide interdisciplinary experience in the academy, we fail to provide the talent necessary to overcome the important, domain-specific issues that are endemic to every industry, country, and profession. Social science graduates – many of whom continue to school in law, business, and other disciplines – may never get another opportunity to understand the power of data science in their disciplines. This places these graduates at a significant disadvantage and hobbles the contributions they might offer. Students need learning opportunities that bridge data science theory and methods with politics, economics, and business to flourish in today’s professional environment.
John Tukey’s philosophy was that the method is never as important as the actionable insights generated with it. I am a Tukey adherent. My courses apply this philosophy to build the marketable skills and purpose students need to take leadership roles in the professional vanguard. These courses anchor in substantive research applications that require deeper consideration of the methods used to develop them. Students must blend comprehension of the manifold capabilities – grounding in theory, analytical techniques, applications to value creation, operating models, standard processes, people management, architectural paradigms, ethical concerns – to succeed in the quantitative sciences, and my courses hold true to that. My courses require students to integrate across these areas to produce a singular result, and they are evaluated on their ability to produce an end-to-end work product.
This philosophy inspires the curriculum I’ve developed for the Quantitative Theory & Methods department at Emory University, and oriented the courses I taught at Columbia University. My university instruction has since 2016 included courses in Data Analytics (Emory, QTM385), Data Science for Startups and Enterprise (Emory, QTM285/385), Electoral Forecasting (Emory, QTM285/385), Data Analysis & Statistics (Columbia, POLSW3704), The American Congress (Columbia, POLSW3222), and Introduction to American Politics (Columbia, POLSW1211). The courses have instructed undergraduates and graduate students at all levels. The courses have varied in size and format, from seminar to major-requirement lecture. I have instructed a total of 560 students in courses for university credit. My average rating across these courses and students is 8.8 out of 10 (with a high of 9.6, and low of 8.0).