Big data has made its way into higher education.
Industries ranging from consumer marketing to government intelligence and crime analysis have used the large swaths of personal information for years. Now higher-education leaders have turned to those data sets to tell them something more about their student body.
For years, colleges and universities have collected information from students and student groups in areas including enrollment, retention, financial-aid awards and graduation, but it was mostly for reporting purposes. For the most part it's been "autopsy data," or information on students who are no longer enrolled.
Now colleges and universities are looking at big-data techniques to help students earn a degree, a task that is growing in importance as the state changes its way of funding higher-education institutions from a model based largely on enrollment to one focused on the success of students.
"I expect to hear a lot more of the schools adding these types of data-analytics products to their student information systems for early alerts, especially for those students who are marginalized," said Maria Markham, director of the Arkansas Department of Higher Education. "The new [funding] formula is going to incentivize doing a good job with certain populations -- not just getting them enrolled, but successfully getting them through the pipeline. So since the focus has shifted ... it's very likely to prompt institutions that don't have these types of capabilities to go out and get those capabilities."
In Arkansas, many higher-education leaders are gearing up for the funding method -- which will be before legislators as a bill this spring -- by investing in the emerging field of big data.
It's in part because the information has become easier to collect, said Michael Moore, vice president for academic affairs at the University of Arkansas System. But also the local and national conversation has shifted, homing in on a call for colleges and universities to no longer just say they are successful, but to show it, he added.
That is coupled with the goal set by Arkansas leaders to bring the state's historically low percentage of Arkansans earning a credential -- a technical certificate, associate's degree, bachelor's degree or higher -- from the current 38.8 percent to 60 percent by 2025.
Enter predictive analytics.
Similar to online retailer amazon.com, predictive analytics pulls data from a student's past performance and other students like him to anticipate how the student will do in a certain course or degree program. That can help identify students who may be at-risk early on, and intervene to help them succeed.
The approach works well for students in general, but there could be less of a focus on students who are not considered "high risk," said Jeff Hankins, spokesman for the Arkansas State University System.
[EMAIL ALERTS: Sign up for free Arkansas Democrat-Gazette email updates]
Ideally, that information will be replaced by real-time data on student performance, such as attendance in a class or performance on assignments or exams, Moore said.
Some universities, such as the University of Arkansas at Fayetteville, already have products in place to help students succeed. Others are starting to put something in place, and for many others it's high on the wish list.
"The better and more dynamic the system, the more it's going to cost," Markham said.
The schools also will use the analytic tools in different ways. For example, Moore said, a university looking to raise a low retention rate will look to use data analytics for more broad-based recommendations, but a flagship four-year that aims to raise a retention rate by a few percentage points may look for more specific measures, such as a single course in the core curriculum that is driving students out.
The University of Arkansas at Monticello -- as an open-admissions university, one that allows students to enroll regardless of test scores or high school grades -- gets students from across the spectrum: high-performing ones to others who need developmental coursework. Nearly two-thirds of the 458 students who started last fall needed developmental course work, state data show. And about two out of five students this fall received the federal Pell grant, aid for low-income students that does not need to be repaid.
Chancellor Karla Hughes said administrators have to consider what student success looks like for the spectrum and what it might mean for the institution.
"For us, the idea of outcomes-based funding has introduced a real need to understand and look at the data for our students, which I will have to say was not a common practice on this campus," she said. "We need to look at what data do we need, what our data is telling us and, based on that, where do we need to put our resources."
The university has become a member of the Student Success Collaborative, an umbrella of the Education Advisory Board that helps schools gather student data and format it into dashboards that faculty members and advisers can understand. Students also will have access to their own dashboards, she said.
Through the collaborative, the school will get faculty members trained on how to read the dashboards and make data-driven decisions from them, Hughes said. Those faculty members will teach others at the school.
Administrators are carrying out best practices, such as requiring all first-time students entering to live on campus and allowing juniors and seniors to register for classes before underclassmen. The university is also transitioning its student union into a $7.5 million Student Success Center, a one-stop shop for students to take care of admissions, advising, financial aid and other wraparound services, she said.
Hughes is not worried about the new way to fund public colleges and universities, one that will place a greater value on helping underserved students that the university takes in. UA-Monticello is taking the right steps to get ahead of the new method, she said.
"I have concerns that we can't do PD [professional development] quickly enough," she said. "I know that we will need some additional advisers, some additional infrastructure. We will know basically what the rules are and we will be able, through the efforts we're putting in place, to monitor those.
"Do I expect to turn around our enrollment and retention in our first year? No, it will be an incremental change. It's going in the right direction that's important."
Henderson State University also started ahead of the curve, said Steve Adkison, the school's provost and vice president for academic affairs. Two years ago, the Arkadelphia school started crafting its strategic plan, one with a No. 1 priority of enhancing student success, he said.
Within that priority are five broad strategies to accomplish 16 goals, many of which include increasing retention rates from freshman year all the way to senior year, and ultimately, raising the graduation rate.
Roughly half of Henderson's 3,567 students are the first in their family to go to college, while about 42 percent received federal Pell grants.
Last year, the school hired four professional advisers and assigned them to academic programs, Adkison said. The professional advisers build stronger relationships with the faculty and students. The school also is taking part in the Higher Learning Commission's Persistence and Completion Academy.
The academy is a four-year program that helps institutions evaluate and enhance retention and graduation. It allows participating schools -- accepted in clusters -- to share ideas and strategies that have worked on their campuses.
As a part of that, Henderson is working to improve its information systems and data development, Adkison said.
"Our information systems here on campus are obsolete," he said. "They don't provide the data we need in the 21st century. We've got considerable amounts of data, but we need to integrate those so we have a coherent set that everyone across campus is aware of."
The University of Arkansas at Pine Bluff has had demonstrations from vendors and is reviewing products, said Linda Okiror, the university's associate vice chancellor for enrollment management and student success.
In the meantime, the university has installed a new program called Student Planner Module that helps students map out their courses over multiple semesters -- which can reduce time to graduation and the cost of college -- and will help the school plan logistics, such as how many of a certain course it needs to offer, she said.
UAPB started an early-alert program this semester to catch students who may be struggling. And thanks to alumni, the historically black college is giving out small grants to students who have 90 or more credit hours but need help getting to the finish line, she said.
Okiror said the university is worried about the new funding method, but it is positioning itself to do better.
"When you consider the students who attend UAPB, they're 90 percent Pell-eligible and over 75 percent take remedial coursework," she said. "We are concerned, but we have a lot of things in place. I'm real happy with what we've been putting into place in the last two years and how it's affected our enrollment."
Universities with the data collection and mining tools -- and the companies with which they contract -- still must comply with the Family Educational Rights and Privacy Act, a federal law that protects student records. Companies usually sign data integrity agreements to ensure adherence to the Family Educational Rights and Privacy Act.
Markham, the Higher Education Department director, said the tools -- if used effectively -- can be a game changer in getting more Arkansans credentials.
"Big data will help any institution if they have the human capacity to carry out whatever needs to be done," she said. "It has to be something that's very accessible, very easy, very streamlined."
Big data is just another tool in the toolbox, said Moore, the academic-affairs official with the UA System.
"You still have to have the wherewithal to implement solutions," he said. "There is still a difference in my mind between the rich data analytics and that these sophisticated programs are doing and a tool that writes reports. The real cutting-edge stuff is on the predictive analytics."
He cautioned that it's still an emerging field.
"It's not a silver bullet," Moore said. "It's not going to solve our problems. To me, the real promise is helping us find those things that are either barriers to success or things that we, through close observation, don't see that are right before us.
"The real promise of this is helping us design and evaluate interventions in learning."
A Section on 11/29/2016