Predictive analytics driving university practices


Universities are using data to help make more calculated decisions.

As campuses seek to target prospective students and spend precious funds most effectively, more administrators are learning about predictive analytics—statistical techniques that gather data and help campus leaders make predictions about the future.

During its Business Analytics Online Education Conference, IBM tapped a number of college administrators to share their experiences with predictive analytics, and how the trend is helping to improve campus operations.

“Predictive analytics plays a very big role in terms of enrollment,” said Jimmy Jung, assistant vice president for enrollment management at the College at Brockport, which is part of the State University of New York system.

Using historical data, retention and graduation rates, application counts, and demographic data, Brockport analysts are “able to predict earlier and react to an increasingly competitive environment,” which Jung said has resulted in an increase in campus applications.

See also:

Special report: Smarter education

Analytics use boosts student retention

Ed-tech group to push for more analytics use in colleges

Brockport staff also have use predictive analytics to examine the impact that the campus’s current financial aid strategy has on enrollment, in order to objectively decide what is and is not working.

This is Brockport’s first year using predictive analytics to drive decisions, especially where recruitment and enrollment are concerned.

Jung said IBM’s SPSS “has made a drastic impact … specifically targeting markets where there’s a higher potential for growth.” The software has helped Brockport officials determine which students have a higher probability of enrollment, enabling the school to spend more money in those areas to increase that probability.
Brockport created an online early warning system to help flag at-risk students with the use of predictive analytics. Student, faculty, and staff surveys performed at certain points early in the year help to identify students who might need extra help adjusting to a college environment.

Jung said that closely analyzing data “really taught me to look beyond the traditional end-of-term data files that institutions use.”

He predicts that analytics will become more instantaneous and more automated in the next five years, with more integration between systems and programs.

Using SPSS, Wichita State University is creating a unified student course system, said David Wright, the university’s assistant vice president of strategic planning and business intelligence.

“The data system gives business analysts comprehensive data on students, course records, prospects, and degree completion,” Wright said. The system also offers data on financial aid, housing, and other relevant information.

See also:

Special report: Smarter education

Analytics use boosts student retention

Ed-tech group to push for more analytics use in colleges

Wright said the university uses SPSS to identify the best and high-yield student prospects, as well as identifying at-risk students based on information ranging from pre-enrollment statistics, standardized test scores, current academic standing, course completion, and changes in academic behavior including a switch from full- to part-time student status.

Wright said the university’s recruitment, enrollment, and administrative practices have changed as a result of insights using predictive analytics.

“We basically have completely revised the way that we do admissions,” he said. “[We are] truly engaged in evidence-based decision-making using data available through SPSS to make decisions about recruitment, personnel, and vendors. It’s really a paradigm shift from where we were several years ago.”

Using predictive analytics, Wright said university leaders are able to identify which prospective students are most likely to attend the school and which geographic enrollment-related events are best to attend.

The university has seen a “huge cost savings related to publications, mailings, travel, and events,” he added, estimating savings of tens of thousands of dollars.

Concerning enrollment, SPSS helps Wichita identify at-risk students even before classes begin. Examining students’ courses and other data helps university staff ensure that those students are prepared, and “allows us to have higher retention, because we’re able to keep these students for a longer time,” Wright said.

“Data now is first and foremost in a budget discussion,” Wright said of how predictive analytics has changed the university’s administrative policies. “Data is no longer an afterthought. Data is now in the room where strategic planning is discussed.”

While Wichita has used predictive analytics to focus on recruitment and retention, Wright predicts that the practice will expand into the university’s financial practices, including financial auditing and forecasting.

“We can also use SPSS to develop alternative funding models, and these can be tested to information budget allocations before a decision is made—these would affect not only cost, but performance issues,” he said.

See also:

Special report: Smarter education

Analytics use boosts student retention

Ed-tech group to push for more analytics use in colleges

Ellen Wagner, executive director of WCET, said her organization is hoping to examine what a number of institutions are doing with predictive analytics in an effort to identify large patterns and help other make decisions that are informed by university practices. WCET works to advance technology-enhanced teaching and learning practices in higher education.

“People are fascinated at the power and strength of decision-making that comes from predictive analytics when it deals with tangible things that are fairly easy to quantify and measure, i.e. planning from an academic perspective,” she said.

Wagner said WCET is seeing different institutions start to share data and change the way they interact.

“Every institution really does have a unique population that it serves really well. … You have a chance to see where the different types of students are particularly successful; it gives people the opportunity to do institutional selection,” she said. “In the same way that institutions look for certain types of students, the students themselves are looking for the same types of advisements. Being able to identify the places where one will be more likely to be successful than another is a fairly compelling way to begin looking at this issue.”

“I want us to be able to really take advantage of a lot of the different convergence trends,” Wagner said, adding that future applications include “using predictive analytics to make education a better place.”

Sign up for our newsletter

Newsletter: Innovations in K12 Education
By submitting your information, you agree to our Terms & Conditions and Privacy Policy.

Laura Ascione

Oops! We could not locate your form.

Sign up for our newsletter

Newsletter: Innovations in K12 Education
By submitting your information, you agree to our Terms & Conditions and Privacy Policy.