Preparing for a Data Science Future
There are many analogies in sports for predicting the immediate future: anticipate the pass, skate to where the puck is going to be, and beat them to the punch. Even though technology keeps moving our goalposts in higher education, we know where the jobs of the future are going to be: data science and data analytics.
In addition to the focus on data science and data analytics, interdisciplinary collaboration is also vital to success in undergraduate education
If you say machine learning, artificial intelligence, big data, or any other trendy word from a headline or article, most people will pay little attention because it doesn’t feel relevant to their daily life. Even though data science is all around us, you can frighten or annoy people with tech talk because not enough effort is spent on relating data science and data analytics to everyday life. We can change that in higher education.
The rate at which information is produced is exponential, with estimates that 90 percent of all the data in the world was generated in the preceding two years. Current projections are that Internet of Things (IoT) devices will outnumber the world's population for the first time this year. There are more articles about data being the basis for growth for companies, and the U.S. alone faces a shortage of 1.5 million managers with the skills to analyze data to make decisions. Today’s “big data” will become meaningless because there will always be bigger data a few years from now. Given this reality, our students need tools today to work with data within their chosen professions in the future.
In response, an increasing number of universities have launched data science programs, some with new data science institutes or centers. While it is great that there are new and exciting master’s programs available, the applicability of data science and data analytics is broader, and it should not be limited to a single discipline. It should be integrated into all degree programs, and specifically, the undergraduate curriculum.
In higher education, we need to evolve our programs to elegantly use data science without being perceived as a geeks-only degree. At Southern Methodist University (SMU), data science is defined as an interdisciplinary field consisting of methods and systems to extract knowledge and insights from data. It encompasses statistics, machine learning, visualization, business analytics, data analytics, and scientific computing. SMU has launched an interdisciplinary undergraduate program around data science as well as a Ph.D. in data science. Additionally, SMU has started construction on the Ford Research and Innovation Building, an interdisciplinary gathering place that will be home to the gaming program, high performance computing, and data science for students of all disciplines. This type of “all-in” approach is a reflection of the importance of data in all programs of study.
In addition to the focus on data science and data analytics, interdisciplinary collaboration is also vital to success in undergraduate education. The Illinois Institute of Technology’s IPRO program is an innovative undergraduate experience that is about 20 years old. Students are required to take two interdisciplinary courses in order to receive their bachelor’s degree. By making the interdisciplinary courses mandatory, it creates an environment of innovation and engagement for all students, not just those in certain disciplines.
While some programs use terms like ‘interdisciplinary,’ ‘collaboration,’ or ‘inter-professional’ as a way to describe their programs, SMU uses “inquiry” to represent interdisciplinary innovation based in a data-driven world, using creative and interactive technologies. The inquiry program is designed to be an interactive experience, and it features the critical use of data about people, society, basic research, and the creative experience. As an example of socially relevant data analysis, SMU’s National Center for Arts Research conducts data-based research on the economic health of the arts and culture industry covering 2,700 arts organizations across 11 arts disciplines. This demonstrates the shift from “data science” and “liberal arts” as two different worlds to a comprehensive fusion of “data arts and sciences.”
Gaming is also a key component of data science. At New Mexico State University, the gaming expertise is embedded in the college of agriculture, and it is used to facilitate the extension mission by making games for children. Because the gaming program exists, it can partner with other disciplines throughout NMSU to develop creative and interactive ways to teach and perform research. A properly designed game is a tool for understanding complex problems elaborated by users who become co-creators in the act of discovery and visualization.
On the flipside, too much data is bad for the soul, and when students experience information overload, it produces lapses in knowledge and results in a failure to act. Solutions to information overload will come from those who understand Aristotle as well as Python. It will require domain experts in all disciplines to determine how information can be deployed, integrated, analyzed, and experienced in the ways needed by a society struggling with so much data. Regardless of their academic program, students should be able to fearlessly engage data, collaborate on intellectually diverse interdisciplinary teams, embrace the media of the interactive era, and contemplate the impact of innovation and disruption on society.
Given all of this activity, how does a CIO go about supporting it? At SMU, we have one IT department, and it covers all aspects of teaching, research, and administrative IT. Our academic technology support includes a representative from the IT shared service that is embedded in each of the schools to work directly with the faculty and deans. Additionally, our centralized data science support service includes staff specializing in high performance computing, Internet of Things, data science, data analytics, and data warehousing. These specialized and centralized IT staff work alongside our infrastructure, application development, and help desk teams. This combination of centralized and embedded IT staff, that are all part of the same IT unit, allows us to support data science and inquiry initiatives throughout all schools, departments, and programs at SMU.
The IT staff embedded in each of the schools, meet on a regular basis, and they assist in translating local needs to central IT projects. In practice, this means that the same Linux administrators support both the high-performance computing cluster and the ERP servers. Our database administrators and integration teams link software together, create administrative datamarts, and help researchers in the school of education develop datasets for studying learning effectiveness in local school districts. Our IT group teaches R, studies classroom utilization data, and helps faculty change traditional physics experiments into IoT exercises. Being able to cover a wide range of IT problem-solving with the same staff allows us to minimize expenses and maximize the breadth of service.
Data science, data analytics, gaming, and interdisciplinary learning will unlock the jobs of the future, and higher education is fundamental to preparing the next generation of workers who embrace data every day. Being well versed in data does not mean that you have to be a geek, and what we have shown with our IT service model is that you can support big ideas without extraordinary increases in IT funding and without creating unnecessarily redundant IT groups. Such an approach may be your game winning strategy.