08-05-2024
Data science has become integral to both academic and business sectors, fueling crucial discussions on its definition and future directions. These discussions play a key role in shaping organizational strategies for data science investments and talent acquisition.
Earlier this year, the Harvard Data Science Review (HDSR) published an article title Data Science at the Singularity · Issue 6.1, Winter 2024 (mit.edu) by Stanford University professor David Donoho. He introduced the concept of "frictionless reproducibility," emphasizing the value of data science in enhancing research collaboration and leading to seamless sharing of data, code and ideas among researchers.
The article generated significant interest, prompting 16 commentary discussions from various fields, including a piece I wrote titled Rethinking the Data Science Singularity · Issue 6.1, Winter 2024 (mit.edu). The forum concluded with a follow-up discussion by Donoho, Rejoinder to Discussion of "Data Science at the Singularity" · Issue 6.1, Winter 2024 (mit.edu).
A key insight from the forum is the view of data science development, not as a universally agreed-upon journey but as a collection of exploratory travels in the digital world. Academia often acts as a travel agent or service provider, developing and guiding common paths, while businesses are likened to travelers with diverse and specific goals, experiences and resources. For example, the consumer credit scoring industry has made significant strides in sharing data and ideas among stakeholders such as retailers, banks and insurers, whereas other sectors are still building foundational data infrastructure. Another example is that generating reliable key performance indicators (KPIs) may fall under the purview of data scientists in some companies and business analysts or even IT professionals in others.
This perspective suggests that each organization should have its own unique approach to data science. It's never too late to begin the journey, but not starting a data science journey can lead to missing out-of-box learning that often equates to competitive advantages. While academia may set a broad itinerary, businesses should tailor their paths based on specific organizational culture, needs and goals. Moreover, there can be interchangeability in professional titles like applied statistician, business analyst or data scientist across different business contexts, which may not be a norm in academia.
Zhiwei Zhu is a clinical assistant professor of management in Purdue University’s Mitch Daniels School of Business. He serves as academic director of the Business Analytics and Information Management undergraduate program, and was recognized as a “top 50 undergraduate business professor” by the website Poets & Quants in 2023. Prior to joining Purdue, he spent more than 15 years in business analytics functions at major global and U.S. insurance companies. He is the recipient of a 2024 Fulbright U.S. Scholar grant award for international education collaborations.
