Issues in Education

Get out of your comfort zone

By Vijay Mehrotra

In my recent MBA elective course on data mining, I assigned my students a term paper. The assignment required students to clearly describe a problem that they thought would be amenable to machine-learning methods and then discuss how this problem might be solved (what techniques they would suggest using, what data would be required and how that data might be captured, what types of benefits would accrue and to who, what barriers to implementation they might encounter, etc.).

This was the first time I had tried this assignment. As with any experiment, I was not sure what kind of results it would produce. The students chose topics ranging from reasonably frivolous (a recommendation app for which beer to choose in a crowded bar) to energetically pragmatic (how to choose a second market for an organic grocery delivery company and how to predict the likelihood of a video advertisement going viral) to ambitiously altruistic (how to proactively issue an alert when an artificial heart is likely to fail). The quality level was equally varied.

After learning algorithms for classification, prediction and clustering, several students reported that this assignment forced them to be more creative, to think through more complex problems, and to confront data issues at a very different level. “Definitely got me out of my comfort zone,” one of my students told me.

After reading these papers, I still had a nagging suspicion that perhaps all of this effort was largely for naught. Most “real” analytics applications have a great deal of complexity, especially in data preparation and in feature engineering, and it seems naïve to think that my MBAs are truly prepared to tackle them on their own.

As I was chewing on this, my friend and colleague J. D. Schramm pointed me to an insightful blog post [1] co-authored by Ali Chaudhry and Rishabh Bhargava. Motivated by their own experiences, the authors first candidly describe the way in which MBAs and engineering students regularly fail to connect with one another, which in turn had led them to conduct focus groups and distribute surveys to try to better understand this phenomenon. From here, they summarize the results of their analysis and then conclude by making some recommendations about how to improve these cross-campus relationships.

One of their key findings was that engineers often do not see the value that MBAs bring to the table. Reading this brought a wry smile to my face. It has been more than 25 years since I finished graduate school in engineering, and more than a dozen years since I started teaching MBAs. Nevertheless, much like the engineering students that Chaudhry and Bhargava had interviewed, I still seem to have a subconscious sense that MBA skills are “easily learnable,” whereas technical knowledge and skills are fundamentally difficult and out of reach for most mere mortals, I realized that it was probably this bias that was at the heart of why I was lamenting the value of the papers that my MBA students had written for my data mining class.

Scott Hartley would suggest that it is high time we all get over this kind of bias. Hartley, a venture capitalist and start-up advisor, is the author of a recent book entitled, “The Fuzzy and the Techie: How Liberal Arts Will Rule the Digital World” [2]. Hartley convincingly argues that in the digital age it is ever more important to have an understanding of what we want technology to do for us, and why. “We must embrace technology,” Hartley opines, “yet not forget the liberal arts that give meaning and explain why, not just how, we build.” Describing many of today’s (and tomorrow’s) leaders, Hartley asserts they must “know enough about technology to partner with techies to execute their vision. Their comparative advantage is in their ability to identify problems, and ask questions...”

This is exactly what had prompted me to assign my data mining students to write their essays, and more generally why I believe it is valuable to teach analytic thinking and skills to MBAs.

While reading Chaudhry and Bhargava’s blog post and Hartley’s book, I felt myself breathing an involuntary sigh of relief and feeling an unexpected sense of validation, not only for the data mining term paper experiment, but also for my vision for a larger project that I am working on.

Long ago, I lamented that, “The non-mathematical factors that help determine success [of analytics projects] are rarely discussed by our journals, our textbooks and our courses. And that is both a shame and a disservice” [3]. A decade later, after being tenured and promoted to professor and rapidly approaching retirement age, I have finally decided to do something about this.

In 2017, I began writing a book that focuses on the many human interactions that take place while trying to make mathematical models work in innovative ways. The book is intended for students who are either preparing for or already working on real-world analytics projects as part of their academic training, for instructors who are responsible for supporting them on these projects, and for newly minted analytics professionals looking to gain a broader perspective. My goal is to supplement the readers’ own first-hand experiences by providing my own perspective in my own voice, informed by my own career adventures.

The book’s format is a cross between a textbook and a collection of short stories. Much of the content is presented in the form of essays, with many of them built around stories. This is a deliberate choice. This is partly because of my desire to get these messages to stick (as Rudyard Kipling famously said, “If history were taught in the form of stories, it would never be forgotten.”) and partly because the act of telling these kinds of stories can reveal the key ideas and issues more effectively than bullet points and jargon.

As yet, there is no definitive publication date for this book. I managed to get a good start on the project during my sabbatical from teaching last spring, but I still have a lot of work to do. Moreover, while I have a fairly concrete outline, it still feels much of the concrete is still wet, waiting to be spread and shaped. Like my data mining students writing their term papers, I too am definitely out of my comfort zone.

Vijay Mehrotra (vmehrotra@usfca.edu) is a professor in the Department of Business Analytics and Information Systems at the University of San Francisco’s School of Management.

Editor’s note:

An earlier version of this article appeared in Analytics magazine.

References

  1. https://medium.com/non-disclosure/across-the-street-the-disconnect-between-mbas-and-engineering-students-38ab22e95282
  2. https://www.amazon.com/Fuzzy-Techie-Liberal-Digital-World/dp/0544944771
  3. http://analytics-magazine.org/was-it-something-i-said-singing-the-silicon-valley-blues/