20 Dec Molding Data Into A Tool For Student Retention
More data has instantiated in the past two years than in the entire previous history of the human race. This explosion brings more potential for institutions to map and track meaningful data patterns than ever before, heightening capacity to draw evidence-based insights that can predict student performance and drive support.
Marist College’s academic early alert system, for example, mines past student data to predict a student’s success early in a course and facilitate faculty intervention. But, successful data application is not as simple as having access to high volumes of data.
Pruning data for significant and applicable patterns is work itself— and the more data potential there is, the more work it takes to weed through. Marist College did not just have to find and track factors predictive of performance, but those that were also valuable to intervention. Demographic information, for example, might have some predictive value, but it’s not as useful to intervention as tracking interaction with online study tools. And then there’s the quality of the information gathered; are data collection practices reliable and are they conducted in a way that engages students in their progress?
Strategizing is critical. Practitioners must craft data collection and evaluation practices that best serve application, especially when student success is at stake.
Part Of The Solution
You cannot automate your way out of the work to improve student retention. Students of Promise, Pell-eligible and other students who face socio-economic challenges, need a simple support structure. Practitioners of data in education reform need to position ourselves to be most useful for those we serve, not fill our business analytics tools with data that does not effectively drive student success.
Data mining and analytics in higher education are still new, but examples of data working to support student retention exist. Successful models come from institutions leveraging data analytics to inform significant investment in other initiatives. Georgia State University, for instance, developed systems to analyze data about past student performance in ways that help the university recognize when current students might be in academic trouble. The system triggers interventions to assist the students in getting back on track. As a result, Georgia State has more than doubled its advising staff, to about 100, and now meets the national standard of one adviser for 300 students.
Additionally, in 2016 its four-year graduation rate rose five percentage points, and the six-year rate rose 6 points. Analytics alone is not the answer. But, analytics designed to increase the effectiveness and efficiency of additional programming might be a possible solution.
Data As A Tool For Service
Data practice in student retention requires a constant conversation between collection methodology and practical application because of the personal and emotional nature of details. Research and application must inform one another; analyzing a functional system can start with the methods for gathering and storing data.
Predictive indicators that are most accurate for predicting student engagement and learning outcomes must be the focus of an effective retention program. At BridgeEdU, for example, we track over 50 predictive measures, including attendance, transportation, food instability, housing instability, and FAFSA completion that allow us for partnering with the student early in the semester to overcome obstacles impacting the traditional Key Performance Indicators (KPI) of Grades, GPA, and retention. When actively and accurately monitored and flagged, these data points inform predictive modeling for coaching, recruitment, and intervention strategies.
Proactive program design is essential for our coaches to build such strong relationships with students. These relationships are not only vital for intervention but the foundation for quality data gathering itself. Plenty of systems can “predict” students that need support based on their historical demographic information. But a conversational approach proves advantageous in the following ways: students share more details as the student-coach relationship stabilizes and develops trust, real-time updates from regular meetings allow for early intervention and necessary support modifications, and the pressure points for students identify with a mentor, which cultivates self-accountability.
Implementing electronic communication channels, such as texting, with these students is another way that our qualitative data inform proactive, engaged programming. BridgeEdU’s program achieves a 75% retention rate with students in this cohort because it combines the predictive tracking with coaches who foster a relationship with the student. Once that relationship exists and coaches train on best data collection practices, it is just a matter of using the data to drive coaching to drive successful outcomes urgently.
To learn more about data’s role in holistic student retention programs like BridgeEdU, click here.
By Dave Demsky
BridgeEdU Sr. VP of Product Development