Visit Modern Campus

The Past, Present and Future of Big Data in Higher Ed

The EvoLLLution | The Past, Present and Future of Big Data in Higher Ed
The use of analytics in supporting institutional management has come a long way over the past few decades, but in today’s environment Big Data can have a transformational impact on efficiency and effectiveness in the postsecondary space, and the future looks even brighter.

Higher education institutions are under increasing pressure from public and private stakeholders to increase efficiency, effectiveness and output. These pressures stem from motivations to expand access while decreasing costs and increasing the number of graduates to fill positions that fuel economic growth and community prosperity. Higher education governing boards are calling for increased accountability and transparency, and regulatory bodies seek evidence of compliance in an environment of reduced public funding and increased competition for students.[1]

Universities are complex organizations that are awash in disjointed and siloed data yet starved for actionable information. Big Data analytics promises to turn these volumes of complex, often unstructured data into actionable information and it is more than just the amalgamation of large data sets into a data warehouse. More than just large sets of data, Big Data is an emergent suite of technologies that can store and process extremely large volumes of disaggregated data of various types at faster speeds and at cheaper costs than ever before.[2, 3]

The History of Analytics in Higher Ed

Many institutions began experimenting with Big Data as extrapolations of their data warehouses, business intelligence or data analytics capabilities. In the early 1990s, institutions were writing sophisticated reports with standalone data systems. Institutions’ motivations at the time were focused on institutional inputs around admissions, revenue, expenditures, matriculation and capacity. But there was nothing really Big about these early days, and it was analogous to driving forward with only your rear-view mirror.

The early 2000s brought a shift in focus with increased accountability of institutional outputs. Institutions evolved their focus towards more sophisticated monitoring of re-enrollment, academic performance, and financial information to understand student retention, progression and graduation outcomes.[4] More recently, these models have been augmented with indicators of student engagement such as library use, tutoring services, portal activity, co-curricular activity monitoring and dozens of other measures as institutional leaders sought to measure student engagement, predict their success, improve business practices and drive retention strategies. This era of analytics starts to look Big with more data systems integration, large data warehouses, and sophisticated reporting tools capable of producing predictive products.

Today’s Big Data technology is providing prescriptive insights into student success, informing curriculum design, recommending shortest path/successful path to degree attainment and triggering interventions intending to keep students on track. To the delight of stakeholders, some institutions have experienced transformative results through new practices based on Big Data-informed strategies.[5]

The Potential for Big Data Today

There’s more to be gained from Big Data—much more. Beyond these student success examples, there lies another world of opportunities to leverage Big Data to improve the operational efficiencies and effectiveness of our institutions. Rapid technological advancement in computational power, prescriptive analytics, image processing, sensors and beacons, data storage, systems integration tools, and advanced search capabilities among other key advances provide insights into systems performance, process bottlenecks, hidden dependencies, and other user-, event-, and device-based data in near real time.

For example, most IT systems produce huge log files of detailed data on machine events and user activities. However, most logging functions only illuminate the most severe events given that the logs are typically manually reviewed and less frequently correlated among different systems (e.g. servers, storage, applications, network). Big Data technology allows for these log files to be concurrently ingested and analyzed in near real time, alerting staff to IT situations across the enterprise. This greatly improves IT staff efficiencies and effectiveness in resolving problems in an increasing complex ecosystem of networks, cloud and on-premise services.

Big Data can also provide insights into users by building profiles of applications used, information consumption, location data, and related security events. Student utilization of services around campus captured by beacons can illuminate frequencies of service utilization in libraries, activity centers, transportation, help desks, co-curricular events and computer laboratories among many others. Big Data technology can help institutions identify which students are engaging these services, how often, and their sentiments about the institution to gain insights into student behaviors and successful outcomes.

Big Data technology can also be used to combine various business process workflows logs to provide more complete views of how business is accomplished on campus. Combining CRM data with imaging system workflows, registration processes and other business triggers among disparate systems can provide insights into how students naturally engage or navigate a myriad of often disjointed required activities.

Big Data technology can monitor security logs, alerting IT staff to unusual activity about users like successful phishing attacks resulting in logins from remote locations, last-minute direct deposit changes, or the absence of expected activity. The ability to process multiple log files in real time leads to early detection and resolution of problems, thwarted cyber attacks, reduced blame-games among system owners and stakeholders and improved performance and availability of systems.

In the Internet of things (IoT) every device produces data. While each device produces small and infrequent amounts of data, the total amount of data from thousands or tens of thousands of IoT devices on campus will produce staggering amounts disaggregated data. Applying Big Data technology to IoT provides new insights into how our campuses operate and can lead to new economies of savings, efficiencies and effectiveness models.

For example, most environmental control systems on our campuses have a digital interface. Data collected from these interfaces tell us a lot about resource consumption. When this machine data is combined with environmental data, weather, and building utilization schedules, a more complete picture emerges providing facilities managers with new insights into maintaining energy efficiencies moving our campuses closer to climate neutrality. Moreover, facilities management resources can be directed to resolve issues that produce the greatest positive effects based on a wide variety of factors. These are just a few of the practical uses of big data in our higher education IT ecosystems.

Looking to the Future: Big Data and the Future of Institutional Management

The next generation of Big Data technology will certainly incorporate a layer of artificial intelligence (AI) over petabytes of real-time data moving the state of the art beyond merely predictive information to creating immediately actionable knowledge. Combined with IoT, these systems will reduce dependencies on people to monitor, process, analyze, and create action plans to manage facilities, transportation, create or remove access. Future Big Data capabilities will proactively guide students to prescribed high value adaptive services and people based on today’s data and predicted future needs. I envision personal agents for student success (PASS) delivered in real time to students’ mobile devices. These PASS agents will not only guide students through daily success plans and meaningful engagements, they will also provide access to personalized learning content that adapts to students learning preferences and engagement performance.

Institutional leaders will encounter some anticipated challenges as they seek the benefits of Big Data technology to improve efficiency, increase effectiveness and guide students to successful outcomes on our campuses. First, Big Data can mean big changes for IT staff and users who might struggle with rapid change management, mass customization concepts and the adoption of radically new processes that will undoubtedly have implications for their work. This culture of technology agility, prescriptive systems, and intrusive service is the opposite pole from where higher education was just a decade ago. Everyone won’t want to be on the Big Data bus.

Although the cost of technology infrastructure is increasingly affordable, there is significant cost associated with collecting, storing, developing and managing algorithms to mine data. Big Data tools may require substantial capital investments in equipment or require a shift of resources in operational budgets around staffing, training, cloud services and software licensing costs. Just procuring the technology is not the only cost towards realizing benefits. Investments in Big Data technology should be tied to institutional strategic plans with realistic goals, shared responsibility among stakeholders and service providers, and realistic expectations of return on those investments.

Finally, security and privacy issues pose significant challenges to the implementation of Big Data in higher education. CIOs must educate campus leaders and stakeholder to be cognizant of the ethics of data collection with regard to the quality of data, privacy, security and ownership. Access to Big Data without training or contextual awareness of uses can lead to inaccurate conclusions and generalizations. Output from Big Data systems may be filtered from decision makers or at the other end of the spectrum, producing overwhelming amounts of output that isn’t actionable. Current policies, decision making, and data governance structures may not be adequate to address the ethics of Big Data utilization.[6]

The Lasting Impact of Big Data

Big Data has evolved from basic reporting systems reflecting on what happened (inputs) to predictive forecasting (outputs) to prescriptive actions (optimized results). Everything seemingly creates data and petabytes of it are being generated each day in our universities. But a number of questions remain: Will all this data lead to more innovation or will technological complexities, ethical limitations, weak governance, or simple institutional inertia stifle the use of Big Data technology to create new efficiencies and institutional effectiveness outcomes? Moreover, is it possible to know or do too much to help students succeed and ensure prosperity for future generations? Can we be too efficient or effective in the operation of our campuses? Can we be too proactive in dealing with problems?

Time will tell and results will vary, but I am optimistic that we’ll finally realize the student-centered and data-driven decision-making environment we’ve strived for in higher education since those wonderful course roster reports rolled off the computer center printers in the early days of data processing.

– – – –


[1] Hazelkorn, E. (2007). The impact of league tables and ranking systems on higher education decision making. Higher education management and policy, 19(2), 1–24.

[2] Baer, L. & Campbell, J. (2011). From metrics to analytics, reporting to action: Analytics’ role in changing the learning environment. In Game changers: Education and information technologies. Ed. Oblinger. EDUCAUSE. Retrieved from

[3] Yang, L. (2013). Big Data Analytics: What Is the Big Deal? Retrieved from

[4] Baer & Campbell, 2011

[5] Daniel, B. (2015). Big data and analytics in higher education: Opportunities and challenges. British journal of educational technology46(5), 904-920.

[6] Slade, S., Prinsloo, P. (2013). Learning analytics: ethical issues and dilemmas. American behavioral scientist, 57(10), 1509–1528.

Author Perspective: