Predictive Analytics for CCR: How Do You Predict If a Student Is Ready?

Posted by Jeff WatsonO

Educators must embrace sophisticated predictive analytics tools to help the many students who are neither dropout risks nor fully college- or career-ready.


Data from the National Center for Education Statistics indicate that the average four-year graduation rate among students attending American public high schools surpassed 83 percent for the 2014-2015 school year, the highest it has been in years. This progress notwithstanding, young Americans’ postsecondary outcomes are not all roses. For example, according to Georgetown University’s Center on Education and the Workforce, young adults’ participation in the labor force has been steadily declining for decades, and now hovers around where it stood in the early 1970s. As a result, today’s young Americans don’t begin earning the median wage until age 30, whereas in the mid-1980s young Americans were, on average, earning the median wage by age 26. A recent analysis published in the New York Times shows that these economic trends are much worse for black and Hispanic students, especially males. Likewise, Balfanz et al. (2016) note similar disparities in post-secondary completion rates.


Got Your Diploma…Now What?


In short, while more students are graduating from high school on time than ever before, many students are finding themselves ill-prepared for the demands of their postsecondary opportunities — whether in higher education or the labor market.


In the view of author and education policy expert David T. Conley, this demonstrates a failure of the public school system to ensure that graduating students are college- and career-ready. “The challenge is not to simply get students into postsecondary programs, as daunting as that challenge might be in some high schools and communities. It is to prepare them to succeed in those programs,” Conley writes in his book, College and Career Ready: Helping All Students Succeed Beyond High School. “A high school diploma awarded to a student who is not capable of performing successfully in any formal learning program beyond high school amounts to a false promise to its recipient.”


As Conley’s comments highlight, there is a significant difference between a student whose performance isn’t necessarily poor enough to trigger an early warning alarm and a student who exits high school ready to take a productive step forward in their life. The real challenge for educators is pinpointing those students who are neither at risk of dropping out of school nor entirely college- or career-ready.


For many — especially large — school districts, this may seem like an impossible task, but with the emergence of sophisticated predictive analytics algorithms, it’s becoming more feasible with each passing school year.


Building Data Capacity


Efficient and meaningful data integration, reporting, and analysis is an obstacle to building CCR-oriented decision support system. There is potentially a lot of data in play from multiple sources, and it is important that CCR systems offer educators with meaningful, easy-to-use analytics. Educator time is at a premium and should be reserved for high-value activities that impact teaching and learning. Most importantly, we should remember that many students are already out of time; they are behind where they need to be. The key to successful CCR systems in part depends on ensuring teachers have ubiquitous access to easy to use, meaningful analytics that informs teaching and learning decisions at a very local level.


Indeed, like Fayetteville, Arkansas, Superintendent Steven Weber points out, “If students enter high school with a weak foundation, it is unlikely that they will graduate prepared to enter a two-year college or the workforce.” Fortunately, bridging the data-sharing gaps between stakeholders — not only between middle schools and high schools but between different groups (teachers, administrators, counselors, etc.) within a single school — is not prohibitively difficult; it simply takes a concerted effort to build and maintain cooperative cross-group relationships.


In a recent report the What Works Clearinghouse, Rumberger et al. (2017) note that dropout prevention and maximizing college readiness are two sides of the same coin. Furthermore, part of the response should be focused on building communities of students and educators as well as fostering connections between stakeholders to improve students’ and families’ engagement within these communities. One important implication here is that communities of practice will require communal sources of information and data that are meaningful and easily accessed, yet still secure.


A Sophisticated Predictive Model


At Hoonuit, we have carefully designed our CCR system to identify students’ level of need as well as the factors driving their level of college readiness. Our dashboards are easy to interpret and apply to teaching and learning decisions. Furthermore, our team is rapidly evolving our CCR system to include additional predictors, new CCR milestones, and even earlier grade levels. For small districts, and those with limited historical data, we have built a data lake that allows us to securely and efficiently produce predictive outcomes regardless of our your districts’ prior data capacity.


Learn More

Get in touch

Get in touch