As budgets get tighter and expectations get higher, leaders at colleges and universities across the United States are trying to understand how to keep their institutions viable, successful and able to serve students. For many leaders, finding the answer likely starts with improving their use of data. Data analytics and business intelligence remain peripheral at many institutions today, but for institutions that have leveraged these tools the benefits have been significant. In this interview, Javier Miyares shares his insights on the value and impact data can have for senior college and university leaders.
The EvoLLLution (Evo): Before the explosion of big data and analytics gathering and analysis, how were major strategic decisions made at the senior level?
Javier Miyares (JM): The university previously relied on a silo approach to collecting and interpreting data, with functions such as enrollment management, financial aid, student support services and marketing each supplying information to senior leaders, who would in turn make strategic decisions.
In 2012, with UMUC (and all of adult higher education) facing what I termed a “perfect storm” of rising costs, declining state support, a mature market, increasing competition, and further cuts to education spending in the military and corporate sectors, we looked to analytics to help guide the university along the pathway of stability and growth.
We established a data analytics unit by consolidating and augmenting functions that already existed at UMUC. With the availability of technology, inexpensively robust computing power, and massive data storage capabilities, we built an infrastructure to collect and process enormous amounts of data. We also hired experts in data analytics. The team’s first order of business was to identify core questions that we needed to answer to get us back on a positive path.
I was confident that we were ahead of the curve in building a fully independent analytics unit when I read a 2016 opinion piece in Forbes magazine, Big Data’s Coming of Age in Higher Education, which argued that “the application of data-driven decision making has begun to permeate all aspects of campus life and operations,” and that the year, in retrospect, may be seen as “the jumping off point for policies and practices that define higher education in the digital era.”
Evo: How has this process evolved as more data has become available to senior leadership?
JM: It is no longer a luxury for institutional leadership to have data analytics capabilities. The ability to look at data across an organization—in a much more holistic way—in order to gain efficiencies is a necessity. We now have the ability to take data from different sources within an institution and with different attributes, to drill down further, and to come up with solutions to problems. The combination of our platform, the dashboards we have developed, and the expertise of our analysts, has introduced significant efficiencies in reporting while increasing opportunities to facilitate meaningful and strategic conversations across the institution.
Our leaders now depend on the capabilities of our analytics team. In fact, our leadership holds a weekly meeting of representatives from a variety of teams that work with new and prospective students—marketing, military operations, community college outreach, regional site teams and others. The data analysts participate as equals.
With such large and varied bodies of data, it is important to note that our data analysts are not just “number crunchers,” but highly trained professionals who use social sciences, economics, finance, marketing and other subject areas to help them interpret the data and tell a story.
Evo: What are a few examples of decisions and changes made by senior institutional leaders that are supported and improved by the availability of data?
JM: At UMUC, data scientists initially worked together with outside vendors to analyze millions of records and hundreds of behavioral and experiential variables and uncover trends that have helped increase student success—including the ability to predict with greater than 80 percent accuracy on the first day of class whether a student would pass or fail. This allows us to reach out to at-risk students before they get in trouble. We also changed our registration policy when we learned that students who waited until the last minute to enroll were consistently at the highest risk of subsequently withdrawing or failing. UMUC now closes enrollment four days prior to the first class meeting, while a grace period of four days after the first day of class allows a student to withdraw without a financial or academic penalty. The results have been dramatic. Undergraduate completion rates rose 7 percentage points over four years, while retention rates increased by 4 percentage points.
Through the use of data analytics, we are smarter about—and better able to target—our spending, as well. As a result, we were able to increase new student registrations in the U.S. by 20 percent while reducing recruitment expenditures by 20 percent in fall 2014 as the university was recovering from unprecedented enrollment volatility that began in 2012.
Evo: At a purely practical level, what are the benefits of dashboards for institutional leaders?
JM: UMUC has developed and evolved a technology platform that encompasses data from across the university. The platform allows for increased visibility into the connections between student outcomes, tuition revenue and expenses. The data is then presented via a series of intuitive and interactive dashboards.
For example, we can look holistically at the performance of a particular academic program. The dashboards allow us to compare key metrics across programs and to compare against internal and external benchmarks. We can make better decisions about where to focus our resources.
Dashboards are key as we look at all phases of the student lifecycle, from the effectiveness of our marketing efforts to attract students to supporting them as they work their way to a degree.
Evo: Looking at the counterpoint, what are some of the limits of the usefulness of data when it comes to managing an institution as large and diverse as UMUC?
JM: One of the key goals of our data analytics program is to improve our ability to help our students succeed. This presents a challenge for UMUC as an open institution that serves more than 84,000 students each year. Early on in the development of the program, we used data from our learning management system to help us identify students who were at risk of failing. The data pointed to one group in particular—those who had long gaps between the times they were logging into their classes. In response, we designed emails that would automatically be sent to the identified group with an encouraging message and a list of resources to help get them back on track. What we didn’t realize at first was that many of the students who were receiving the emails were in the military (more than half of our students are active-duty personnel or dependents serving overseas) and who were deployed or out on training exercises. They were not in danger of failing, just unable to log in because of their schedules.
One of the members of our leadership team in student recruitment brought this point home when she said that just because you have the data and have identified your problem doesn’t mean you know the solution that will address the problem or make it go away. It is at this point that it is vital to be able to call on analysts with the expertise to interpret the data from a different dimension, or with different attributes, that will help drill down further.
A key limitation also is the lack of expertise and infrastructure at an institution to build a comprehensive data analytics program. We built our program from the ground up, and as a result of its success in helping UMUC achieve operational and academic efficiencies and improvements, we made the decision to spin off our analytics unit into a new company that will offer these services for a fee to other institutions. Just as it continues to do for UMUC, the new company, HelioCampus, integrates data from an institution’s existing source systems, normalizes the data—adjusting values so they can be compared on a common scale—and models the data in a central repository. Just as important as the technology, HelioCampus team members provide ongoing data analysis and storytelling services to institutional stakeholders to identify, support and expand analytics initiatives across all divisions of the institution.
You Might Also Like
