Student Insights Engine
DATA SCIENCE AS A SERVICE
The Difference Precision Makes
Before & After Civitas Learning’s Predictive Modeling
of at-risk students were correctly identified using GPA. The majority were missed.
of at-risk students were correctly identified using Civitas Learning’s predictive models.
Data Science You Can Trust
Not all action analytics are created equal. Watch and learn in this interview with our Chief Data Scientist, David Kil, as he explores parallels between analytics in healthcare and education.
As David will tell you, the “three P’s” make all the difference in the world. Here’s where our Student Insights Engine (SIE) really shines.
The Student Insights Engine (SIE) uniquely tunes insights to each institution and every student. One size does not fit all.
Our predictions are unique to each institution. They help you see around corners and into the future, continuously matching actions to outcomes.
Our predictions are rigorous where others are not. Our adaptive models update nightly, ensuring continued validity and accurate predictions every day.
Spotlight: Derived Variables
Derived variables are the stars of our show.
Think of each variable as a unit of insight.
Example #1: Degree-Program Alignment Score
This derived variable considers how well aligned courses taken by a student are with those historically taken by peers who’ve graduated with the student’s declared degree. This critical insight enables institutional team members to identify students who may be unsure of their path and take action to mitigate excess credit accumulation.
Example #2: Relative Discussion Board Activity
This variable considers each student’s engagement within the LMS compared to that of their classmates, enabling faculty to identify struggling students and take steps to increase engagement in their class.
Our promise to you is this: the SIE is not a black box.
Our approach is explicit, and our process is transparent.
You’ll never be asked to take our word for it.
What really makes
Before Civitas Learning, many of our partner institutions used common practices like Prior-Term GPA, Academic Standing, Attendance, and Financial Status to determine which students were at risk. They also followed industry learnings, such as prioritizing outreach and intervention for students with low GPAs. These practices are based on “assumptive” analytics.
Higher achieving students who do not meet these at-risk criteria are expected to persist, even though a surprisingly high percentage of attrition still comes from this group.
With Civitas Learning, one of our partner institutions uncovered a previously unidentified group of at-risk students hiding in plain sight.
Our predictive models accurately identified 660 non-resident, first-year female students with a GPA of 2.4-3.0 that were at risk of not persisting, despite good standing with the institution.
Armed with this insight, our partner was able to take preventative action that helped these students before it was too late.
Know Students’ Needs Today
Dozens of factors combine to impact whether individual students will learn well and finish strong. Some of them are intuitive, but many are not.
Importantly, the SIE refreshes data every night and re-scores each student so that you’re making decisions based on each student’s need and risk as of today…not yesterday, last week or last year.
Data tells us each day how students are doing and what they need most. Are you listening to their story?