Big Data Versus Small Colleges
Why is Big Data such a Big Deal?
Institutional researchers at colleges and universities have been collecting and analyzing data about student recruitment, retention and graduation rates for decades. In recent years, however, these activities—under new rubrics such as “Big Data mining” and “predictive analytics”—have taken center stage in higher education thanks to the conviction that today’s technologies can allow us to base critical strategic decisions on solid data models rather than on anecdotal information or personal intuition. Like many private sector organizations that have made effective use of Big Data to achieve competitive advantage, universities now look to Big Data to ensure student success as well as their own future viability.
Big Data doesn’t come easy, however. Large universities that have been leaders in the use of Big Data recognize that the requirements for embarking on this type of strategy include appropriate data governance policies, advanced data management tools, and sufficient staff expertise.
Is it realistic to expect that smaller colleges, especially those with highly constrained resources, can dive into Big Data analytics in the way that large universities do? Probably not. Whether you are plowing through half a million data points or a hundred million data points, you still need to have the right data policies, technical tools and staff expertise to draw upon.
So what can smaller colleges do? Well, first the good news: When it comes to data governance, smaller colleges are generally in a far better position than large universities because information technology and institutional databases are administered centrally rather than being distributed across numerous colleges and other administrative units. This makes it considerably easier to bring the necessary data stewards together and achieve consensus on data definitions, processing rules, reporting standards and other elements of data governance. Political boundaries tend to be lower and easier to overcome. And one of the toughest hurdles—getting institutional leadership to collaborate directly with data analysts to define strategic questions that can actually be answered by available data—is far easier to overcome in small colleges than in large, complex universities.
Now for the not-so-good news: Traditional enterprise databases that store the information required for Big Data analytics are essentially transactional systems. That is to say, they reflect the latest state of each record but do not preserve past states in a manner that allows for trend analysis.
In order to capture and maintain a rich set of past record states, universities deploy data warehouses that are designed expressly for data analytics. Whether onsite or in the cloud, however, data warehouses are costly to implement and maintain and may exceed the resource limitations of smaller colleges. Fortunately, there are ways to address this issue. Data Cookbook, for example, provides a range of warehouse/analytic services especially suited to colleges on tight budgets. Alternatively, newer enterprise systems are designed with data warehouse features already baked into them. And there is always the option of taking periodic “snapshots” of key data points to examine changes over time (though you may need a crystal ball to determine exactly which elements you’ll want to look at in the future…).
Lastly, there is the issue of staff expertise. In recent years, large universities have hired numerous business intelligence (BI) analysts to pump their data warehouses for answers to strategic questions. These specialists, who come armed with a combination of data analysis, modeling and presentation skills, often provide the communication link between information technologists, data stewards and institutional decision-makers. Smaller colleges may be hard-pressed to compete for qualified BI analysts, especially since there is a growing demand for them in every economic sector, not just higher ed. Outsourcing BI analysis is a strategy that some schools are pursuing and various companies, like Civitas Learning, offer a range of analytic services focused on student success. Perhaps the most cost-effective approach in smaller colleges, however, is to leverage existing staff by providing support for them to enhance their data analytics skills. Obviously, “growing your own” won’t work everywhere but a careful evaluation of internal potential is well worth the effort.
It’s a Moving Target
The greatest challenge to the use of Big Data, regardless of whether you are a large university or a small college, is being able to collect the “right” data. While the amount of data we collect each year grows astronomically, much of it is irrelevant to answering key strategic questions. And the data elements that would be of greatest value may be overlooked entirely. Evaluating data collection mechanisms and continuously adjusting them to reflect potential future needs is vital to the success of predictive analytics. We need to be ultra-sensitive to the risk of solving yesterday’s problems rather than tomorrow’s. Once again, smaller colleges with more highly defined student populations, may have an easier time than large universities when it comes to collecting the data needed to guide institutional decisions.
Thanks to small faculty-to-student ratios, professors and administrators are able to make quick judgment calls about their students’ weaknesses or points of trouble—lack of participation in class, fear of making eye contact, the tremors in the voice hiding the embarrassment of being overwhelmed—and act on those observations. But “you’re probably not going to get that same personalized experience” at larger campuses where students are likelier to be the first in their families to vie for a bachelor’s degree and may not know how to navigate both the bureaucracies and expectations of a college education, Milliron said.
Author Perspective: Administrator