Improve student retention and graduation rates: A top priority

Public Sector Technology Writer

Unfortunately, the problems in the U.S. education system are more widely discussed than any successes. However, there are indeed successes, including strides made to improve student retention at multiple levels. U.S. students are graduating high school at a higher rate than ever. The nation achieved a high-school graduation rate of 81 percent in the 2012 to 2013 school year, according to the U.S. Department of Education, the highest rate in five years. One helpful factor is the increased use of data analytics to identify students at risk of dropping out. With that knowledge, schools can take steps to improve student retention and keep kids on track to graduate.

In a recent interview, North Carolina's state superintendent of public instruction, June Atkinson, shared details about how the state leveraged big data analytics to increase high school graduation rates. North Carolina went from a four-year cohort graduation rate of 68.3 percent in 2006 to 82.5 percent in 2013.

Adopting graduation resiliency software

North Carolina developed a solution that analyzed data from its statewide student information system and found that dropout rates were highest among ninth graders. This analysis helped determine which new high-school pupils had the greatest risk of leaving without graduating. It analyzed how individual students' data mapped to known characteristics of at-risk students, including low attendance rates and failing grades in middle-school math and English."Instead of spending a lot of time trying to figure out which students were at risk, we could spend time on helping those students," Atkinson explains. For instance, students could be redirected to schools' "ninth grade academies," which give at-risk kids more attention.

Use quality data to inform analysis

To determine which at-risk indicators should be examined by the graduation resiliency software, Atkinson relied on research from Johns Hopkins University. Their data pointed to characteristics such as bad grades, repeated absences and unsatisfactory behavior, she says.

Other school systems also rely on data about absences, discipline referrals and academic status. Gwinnett County Public Schools in Atlanta learned that the best predictor of graduation was whether a student had successfully completed Algebra I. Academic data also revealed that students who mastered eighth-grade creative writing were more likely to succeed in Algebra I. That insight led Gwinnett to prioritize writing assistance. "As writing course success rates improved, so did Algebra I success rates — and, eventually, so did high school graduation rates," according to Forbes.

Similarly, Miami Carol City Senior High in Florida also instituted programs to examine attendance, behavior and performance data, according to NPR. One in three students flagged for missing school during the 2012-2013 academic year were able to get back on track toward graduating. Among students with behavioral problems, two-thirds made a turnaround.

Present findings in an easy-to-comprehend, actionable format

In North Carolina, each district superintendent is supplied information on annual graduation rates by subgroup, including race, ethnicity and gender. In the documentation, the subgroups are marked with color codes to clearly denote how subgroups in one district compare to others in the state. Red means that a particular group is performing below average in a district compared to their peers in other locations, and green means the subgroup is performing above average.

"That helps school districts target certain subgroups which may have been underserved," Atkinson says. Superintendents also use the charts to determine which districts are succeeding in areas where their schools need to improve. "If they have a problem, they can contact peers to learn how they approach subgroups' learning challenges."

School systems that want to improve student retention should try to pull more value out of core data from student information systems. By analyzing that data to flag any known at-risk characteristics, and by presenting high-levle findings to leaders in an easily understandable format, schools can better target those that need help.

To join the conversation on how analytics can help to stem student attrition and to see how colleges and universities like yours are using IBM Analytics for student retention, visit the IBM booth #1734 at EDUCAUSE 2015 or new solution website, IBM Analytics for Education.