How Data Science Is Improving Higher Education

Increasingly, colleges and universities, as well as governments, are using data science to improve the ways educational institutions do everything from recruiting to engaging with students to budgeting.


Big data represents a huge opportunity for higher education. Increasingly, colleges and universities, as well as governments, are using data science to improve the ways educational institutions do everything from recruiting to engaging with students to budgeting.

Here are six ways data science is improving higher ed now.


1. Increasing Graduation Rates

A primary goal of nearly all universities is increasing their graduation rates. There are many strategies for doing this, and data can help schools determine which strategy will work best for them.

Analyzing data about students who dropped out versus those who completed their programs can help schools identify characteristics of those most likely to leave. It can also enable them to offer resources to those students that reduce their risk of dropping out.

Temple University in Philadelphia has been using data to increase its graduation rate. It found students that were moderately low-income and received partial Pell Grants were most likely to drop out, even compared the lowest-income students who received full Pell Grants. It also discovered that students who had four years of a foreign language in high school were less likely to drop out.

Using that data, Temple provided financial aid to at-risk students and also launched initiatives to help improve students' sense of belonging on campus. The school has since seen its four-year graduation rates climb from 20 to 44 percent and its six-year rates increase from 59 to 70 percent.


2. Improving Recruitment Strategies

Data can also help schools get students to enroll in and attend college in the first place. Increasingly, schools are using data to decide which prospective students to target in order to develop their promotional campaigns and personalize their marketing content.

To find potential students, many schools analyze the characteristics of those from the current year and then target high school pupils with similar qualities. Schools can also analyze data on which students graduated and were most successful to determine the types of people that may be the best fit for the university.

Saint Louis University, for example, realized it had to expand its geographic reach based on a projected drop in the number of high school graduates in the region its students typically come from. It used information from the questionnaires on college preferences high school students complete when they register for the SAT or the ACT.

It looked for students who would be a good fit for the university and indicated they were willing to travel to attend school.


3. Optimizing Spending

When schools track their spending and make an effort to calculate the return of their various investments, they generate a multitude of data points they can use to optimize spending. By combining data from multiple schools, the National Study of Instructional Costs and Productivity, also called the Delaware Cost Study, helps schools discover even more valuable insights.

Around 700 colleges and universities participate in the Delaware Cost Study, which enables schools to compare how much it costs them to deliver one hour of instruction with the costs at other institutions. It also provides data on other aspects, such as teaching loads and how much instruction is done by regular faculty.

The Delaware Cost Study program allows participating institutions to define their peer groups to get the comparisons that are most relevant to their goals. The ability to access personalized data in this clearly defined way makes the information more useful and actionable. Educational institutions can use this data to optimize their costs and increase competitiveness with other universities.


4. Improving Campus Safety

Placing more cameras around campus may help deter would-be criminals and enhance campus safety, but it's also time-consuming to analyze video footage manually. Integrating data analytics systems into security systems, however, can make it much easier and improve safety in other ways as well.

An analytics-enabled security system, such as the one at the University of Tulsa, can define search terms, like the clothing worn by a suspect. The system will automatically begin pulling data that matches those terms. The security team can also set up alerts for higher-risk situations so they know which cameras to monitor at what times.

For example, if a car pulls into a parking lot at night, the security team would get an alert that tells them which monitor to check. Then, if they spot anything suspicious, they can send a team much more quickly.


5. Closing Equity Gaps

Data science can also help academic institutions improve outcomes for groups that often face unique challenges in college, including minority, lower-income and first-generation students. Over the past few years, Georgia State has become a leader in this area, thanks in part to its use of data. More black students now graduate with bachelor's degrees from Georgia State than from any other U.S. nonprofit school.

The secret to the school's success is a data-backed advising system. This program uses an algorithm to monitor student performance. If the system detects a student's work is beginning to suffer, their advisor will reach out to them to provide assistance. The school also offers small grants to students who find themselves struggling to pay tuition.


6. Increasing Student Engagement

Students who are more engaged with their school and campus are more likely to graduate in time and will probably have a better college experience and perception of their university. Tracking indicators of engagement allow schools to reach out to those who have not been engaging.

Nottingham Trent University, for example, tracks card swipes into buildings, library use, virtual learning environment use and the electronic submission of coursework. The school discovered that three-quarters of students with low engagement based on their metrics did not make typical progress from their first to their second year of school.

Based on these findings, the school developed a program in which tutors reach out to students who have had no engagement for two weeks. This helps to move the students in the right direction before low engagement starts to have a more severe impact.

From improving outcomes for students to helping schools be more competitive, data science is playing a growing role in higher education. The trend of using data to improve education isn't likely to slow anytime soon.

Bio: Kayla Matthews discusses technology and big data on publications like The Week, The Data Center Journal and VentureBeat, and has been writing for more than five years. To read more posts from Kayla, subscribe to her blog Productivity Bytes.