Teaching data analytics has been a passion for years, and I find it both challenging and rewarding to make abstract ideas concrete and engaging. Analytics topics require serious hard work. Students, even the brightest and the most committed, need to be motivated properly. My coursework tends to be full of mathematical equations, or abstract symbols. To make these topics easier to understand, remember, and recall, I have designed many activities both inside and outside the classroom to make learning effective. I focus on two key strategies: in-class engagement, and feedback-centric design.  I engage students in class by designing props, hands-on activities, and exercises that bring abstract ideas to life. Outside the classroom, I continue to engage students through what I call “feedback-centric design,” to be elaborated below. First I illustrate my overall approach with examples of how I teach with chocolate.

In-class Engagement with Chocolate

My favorite teaching prop is chocolate. There are so many ways to make abstract ideas concrete using chocolate. For example, when I taught normalization in database management, I always asked students to set up candy shops and update transactions in a poorly designed database. These exercises helped students physically experience the headaches of bad database design, and discover logical ways to improve the design, all the while enjoying the chocolate treats they were trading. See more details on this candy database exercise here.

Chocolate jars are great for demonstrating sampling, binomial distribution, Gini impurity index, and many other mathy concepts. For example, students are always very happy to compute variance of a binomial variable, or the Gini Impurity Index, using the chocolate jars illustrated below.

When teaching the decision tree algorithms, I would bring the following chocolate tree to the classroom so the entire class can split the nodes and compute each node’s Gini Impurity Index manually together.

Although chocolate is my favorite prop, I also enjoy teaching with real estate data and have developed a series of in-class activities using data collected from Zillow that students have always enjoyed. I have presented these activities in the teaching track of the SIGDSA symposium in 2021, with a slightly shorter version available here.

Feedback-Centric Design

After students leave the classroom, I continue to engage them through what I call “feedback-centric design.” Just like machine learning, humans learn through continuous improvement from feedback and error. A big part of my teaching work focuses on the design of labs and assignments. I use a range of automation tools to help improve my abilities to provide frequent assignments and timely feedback to students. Business analytics work requires not only repeated practices, but also contextualized feedback that promotes mental model revision and refinement over time. To minimize opportunities for cheating, I create my own labs and assignments from scratch. I also vary them both within and between semesters.

To summarize, my teaching philosophy is all about engaging the student, both inside and outside the classroom. I spend a lot of my creative efforts on crafting physical and visual ways to help students understand otherwise abstract concepts. I carefully design assignments that challenge and motivate students over time. The goal has always been to lower barriers, increase interest, and deepen understanding.