One of the greatest struggles a school, no matter its size, can encounter is how best to counsel its students. Student counselors are given huge tasks, from helping students graduate to treating them for mental disorders. These tasks can be overwhelming when the ratio between students and counselor seems insurmountable and the tools at counselors’ disposal are limited.
For many higher education institutions, one of the tools being used to lessen the burden is predictive analytics. The collection of student data, whether via computer or pen and paper, has always raised concerns regarding student privacy, the marginalization of certain student groups, and even a lack of transparency. All these factors tend to be magnified the larger the student group grows.
Georgia State University, however, has found a way to use predictive analytics to improve a number of factors for its 24,000 undergraduate students. Looking at how GSU has done so may prove to be a great model for other higher ed institutions as well as secondary schools.
Starting with a Question
GSU’s issue was a lack of resources needed to help students navigate their way to a degree. With many first-generation college students among the 24,000 undergrads, GSU wanted to help improve graduation rates, but there weren’t enough resources. Therefore, the college asked itself if it could apply predictive analytics to its student data in order to optimize counselor and adviser time and raise their graduation rates.
The use of analytics in higher ed is now recognized as the future of education technology. However, analytics can’t be used in a vacuum. If they are, they will most likely lead to bigger gaps than they intended to plug. In order to avoid this, GSU worked carefully to identify exactly what it wanted its predictive analytics to measure.
The university and an outside consulting firm went through the data for 2.5 million grades earned by students over a seven-year period. This analysis resulted in a list of contributing factors that influenced a student’s chances for graduation. For a university with a high minority population, graduation success was imperative, since minorities continue to complete college at lower rates than other ethnicities.
GSU and its consultant created a system that sent early-warning signals not just to the students but to their advisors. The system, a hybrid of both a learning management system and adaptive learning software, two major forces in educational progression, uses more than 700 red flags to identify when a student is struggling to meet graduation requirements.
While the system alerted counselors and advisors when a grade was too low to meet a graduation requirement, or a student didn’t take a required course within the allotted time, GSU recognized that it needed more than tech.
Add the Resources
Before the project was implemented, GSU’s student-to-advisor ratio was 700-to-1. As noted by University of Nevada, Reno, social work professor Mary Hylton, being under-resourced is one of the greatest challenges facing social workers, many of whom are employed by educational institutions. “Social workers are put in the position of tackling the outcomes of huge degrees of social and economic inequality without the resources to always do it adequately,” Hylton stated.
To that end, GSU hired extra advisors and counselors, bringing its ratio down to 300-to-1. It also started to rely more on the advisors to start conversations about graduation, requirements, and gaps for students. Historically, most academic advising and counseling has been initiated by the student.
GSU’s early warning system, called Graduation and Progression Success, or GPS, puts the onus on both the student and the advisor to start a conversation about struggles. Aptly named GPS, the system helps students navigate college on a straighter path to graduation.
GPS has been in place since 2014, and its graduation rates have increased by six percentage points. Students are also graduating in more timely fashion, spending more time in required courses than flailing about, taking coursework that does not help them in their majors. Even within its first year of implementation, GPS helped increase GSU’s total number of students on track to finish a four-year degree on time by 9 percent.
A four-year public university has a number of resources at its disposal that a secondary institution may not. However, this model could be a great blueprint for schools hoping to use predictive analytics to improve graduation rates. If implemented properly, it could help more students graduate high school and go on to college without many of the first-year fears. The key will be having the advisory sources at students’ disposal.