3 Ways Predictive Analytics Can Improve Learner Outcomes
It’s common to have trouble getting actionable insights from your organization’s data. That’s because much of the data you likely are looking at in your day-to-day operations is descriptive; it shows you what has already happened in the past. Think test scores, attendance records, and surveys. The problem is, you can’t change what’s already happened. That’s where predictive analytics come in.
Instead of showing past performance, predictive analytics gives you a glimpse into the future. They show you what’s likely to happen based on the historical data you already own. The way predictive analytics work can be complex—incorporating machine learning, statistical analysis, and algorithms. However the practical ways predictive analytics improve student outcomes are unmistakable.
Improve Graduation Rates
Attention is often focused on students who are already failing. Predictive analytics makes it easier to identify students who are at-risk of failing before it happens; giving educators a better chance of catching them in the learning process and changing that outcome.
One oft-cited example is Georgia State, where predictive analytics have been used to improve graduation numbers since 2011. Predictive data revealed that students who graduated from the nursing program typically earned a B or better in introductory English, whereas those earning a B- often dropped out of the program down the line. This encouraged the school to identify and reach out to nursing students likely to perform poorly in the nursing program; intervening by steering them in a different direction to a program where they were more likely to succeed. Read more about how the Panthers have increased graduation numbers in the last decade.
Decrease Educator Turnover
There is a body of research showing that a high rate of teacher turnover negatively affects student achievements. Plus, high turnover can signal other issues within an educational system that is perceived to correlate with lower student achievement. Educational institutions can learn from for-profit businesses that use predictive analytics to identify teachers who are considering resigning, before it’s too late. Data can’t reveal how a specific teacher feels about their job, but predictive analytics can identify unique markers that signify whether an employee is at risk of leaving. For example, a 2018 study found that schedule flexibility, number of students, and affective commitment to the organisation are a few of the predictors of online teachers’ likelihood of turnover.
What about satisfaction surveys, won’t those identify dissatisfied teachers? In short, no. Again, surveys are descriptive, they tell you what’s already happened. That’s also if the surveyed employee is completely honest, which isn’t always the case.
Increase Student Retention
Students can’t achieve better outcomes if they fail to actually stay in school. Higher Education institutions lose up to a third of enrollees due to “summer melt”— the phenomenon of students set to attend college upon high school graduation but who never actually enroll and make it on campus in the fall. Predictive analytics can help identify students most likely to fall victim to summer melt and enable educational institutions to adapt accordingly. The Florida Institute of Technology for example used predictive analytics to better target their recruitment and marketing, while Taylor University in Upland, Ind. used predictive analytics to determine what students were most likely to enroll once accepted, and focused their incoming student efforts on those more likely to actually be able to join the student body.