Learning Analytics at HBS Online

Data sources (e.g., Salesforce, CP MySQL) pulled together via Stitch and copied into Snowflake for reporting and analysis.

On Thursday, November 17, our colleague Brent Benson spoke to over sixty CoDA members  about the data infrastructure, tools, and techniques used by HBS Online. HBS Online is tasked with bringing the HBS case method, rooted in the idea that students learn best from each other through active and engaged discussion of real-world situations, to learners all around the world participating asynchronously via an online platform. Brent spoke about how viewing data as a strategic asset in this process compelled HBS Online to build systems to collect and effectively use data to make informed decisions.

HBS Online uses the tools Stitch, Snowflake, and GitLab to create a data warehouse that can efficiently report across different kinds of data. This architecture allows them to use their data to answer different kinds of learning analytics questions, such as whether procrastination on an assignment affects performance (it does!), and even enables them to use natural language processing techniques to do sentiment analysis on survey data or grade short answer responses to assignment questions. Brent also talked about some exciting potential next steps for HBS Online, including collaboration with other analysts around the school to better understand our students and programs.

View the full recording of the event (contact coda@hbs.edu for the password) or check out Brent’s slides for more details!