By Susan Grajek, Vice President for data, research and analytics, EDUCAUSE.
Big data and analytics are reshaping everything. Industry is using them to great effect, to better understand markets and customers, manage supply chains, and increase profits. Personalized medicine, fueled by analytics applied to big data, is poised to revolutionize healthcare. Higher education lags several paces behind these fields. Some institutions are demonstrating improvements in retention and degree completion, but most are still using data to monitor student outcomes and activities rather than predict or proactively intervene.
Certainly, trends related to analytics and data are influencing institutional IT strategy, more so than other types of trends EDUCAUSE tracks, including those related to teaching and learning and security and risk1. Data-driven decision-making, enterprise data management, and data integration issues are all already incorporated into or exerting a major influence on emerging IT strategy in at least half of US colleges and universities. Personalized learning, however, is only this influential at one in five institutions.
The promise is there. Now, how to achieve it? How do we extend effective uses of analytics move beyond the early adopters?
« Big data and analytics can transform education, student success, and institutional decision-making in ways that enhance and extend colleges’ and universities’ particular cultures, missions, and strategies »
Higher education institutions need to develop the capabilities to use analytics to benefit students. Most of these capabilities extend beyond the IT department and have nothing to do with technology. EDUCAUSE has developed assessments to help institutions understand which capabilities they need, benchmark themselves against other institutions and aspirational targets, and receive advice for improving.
The EDUCAUSE assessments reinforce the fact that technology plays only a supporting role in institutional readiness. The assessments are based on maturity indices for student success and analytics. They reflect the facts that people, process, and data are just as important as technology. Components of maturity include institutional commitment in the form of leadership support and adequate funding; good data; collaboration across IT, institutional research, and constituents (student advising, faculty, students, etc.); clear processes and policies; and, yes, technology.
Student success analytics require new technology investments. They include data management technologies to collect, integrate, and manage data from disparate sources over time; analytics technologies to analyze and model data; visualization and reporting technologies to enable constituents to view and interact with data; and specialized applications to support components of student success, like academic early alerts, degree planning, course recommendations, and advising center management.
« Analytics applied to data from wearables offer the potential for learning more about people’s behavior, particularly if they begin to automatically interact with institutional applications »
The level of investment is comparable to an enterprise resource planning (ERP) project. Yes, that big. But ERPs, while mission critical, are focused on institution’s business functions like HR, finance, and payroll. Big data and analytics can transform education, student success, and institutional decision-making in ways that enhance and extend colleges’ and universities’ particular cultures, missions, and strategies.
To achieve this promise, start with the basics: people, process, data, and technology.
Have you mastered the basics? Then give thought to the Internet of Things (IoT). The number of computers and servers connected to the Internet is being dwarfed by the number of other physical objects with embedded Internet-capable technology. Gartner estimates that the IoT will encompass more than 20 billion devices by 2020, a fourfold increase from 2015. Two-thirds of those devices will be consumer-level devices. This enormous change will increase bandwidth needs, contribute to privacy and security challenges, introduce new computation needs, and potentially provide enormous opportunities for institutions.
For higher education, perhaps the most obvious opportunities initially will be in automating and enhancing infrastructure management. But analytics applied to data from wearables and other person-based devices offer the potential for learning more about people’s behavior, particularly if they begin to automatically interact with institutional applications. Considerable opportunities may present themselves for researchers, particularly in biomedicine and social sciences. No “killer thing” has surfaced for teaching and learning. Yet.