We are pleased to announce our first issue of Community Insights – a deep data research project revealing student success trends and emerging benchmarks from across our partner community. In creating this report we aggregated intelligence across 2 million active student records to explore the most meaningful trends, not just the most conventional best practices. Core to Civitas Learning’s solutions is the ability to glean insights, then take action based on each institution’s unique data and needs. But, there is also great opportunity for this growing community to learn from emerging data-informed signals across the broader ecosystem. The research illuminates some new opportunities to drive improvement, often dispelling myths and conventional wisdom, as we all get closer to our goals for student success.
Our first issue of Community Insights dives into the data and shares knowledge around these findings:
LMS engagement data is highly predictive of student success – regardless of institution type.
More students are leaving our institutions with GPAs above 2.0 than below.
Course grades matter – and it’s not just failing grades. Learn about courses where only an A increases a student’s likelihood to succeed.
It is now possible to alleviate initiative fatigue by determining which initiatives are working and which aren’t. See what Illume Impact is revealing about programs and their ROI.
About the Community Insights Report
Community Insights is based on collective data from across the Civitas Learning customer base. This particular study included 55 institutions and 2 million active students. The data was analyzed using Civitas Learning’s Student Insights Engine and all of the modules in the Illume® application, including Illume Students, Courses and Impact. Illume empowers institutions to use predictors and sophisticated, but easy to use filters to better understand their students so the actions, interventions and outreach they provide students are more successful.
What follows is a glance at the first benchmark. Learn more by downloading the report and watch a recording of our webinar.
Community Insights Benchmark #1: Engagement Matters
What the Data are Saying
We found that Learning Management System (LMS) data are highly predictive of success across most institutions in the sample, including traditional institutions that offer most of their courses on-ground. The type of activity that was most predictive, however, varied by institution. For example, for one it was discussion boards and for another it was engagement in course content that proved most predictive. The way institutions used their LMS varied, but the more they used it within and across courses the more predictive the data was of student success.
Partner Insight: Early Term LMS Engagement Greatly Impacts Persistence
In zooming into one partner’s data we found this four-year, research, on-ground institution’s number one predictor for persistence for first-year students was LMS activity. Specifically, the percentage of days students logged into the LMS in the first 14 days of the term was most predictive of their potential to persist. The average persistence rate for these students was 88 percent. When they logged in fewer than five days of the first 14 days their persistence dropped to 76 percent compared to up to 92 percent for students with five or more days of activity.
The Community Insights report highlights how this institution honed their outreach to specific, at-risk students by using Illume Student Lists.
Benchmarking the Civitas
This benchmark is based on 23 institutions including 12 on-ground and 11 online colleges and universities.
Overall, 80 percent of the institutions had LMS activity in their top 10 predictive variables.
For first-year students, 20 percent of the top 10 predictors were engagement data.
LMS data was strongly predictive for both on-ground and online institutions.
64 percent of the shifting we saw in predictions during the term was due to LMS activity.
Why it Matters
While LMS activity and engagement are highly predictive, understanding which specific activity is the most predictive at any institution requires a closer look into how they specifically use their LMS. What is predictive at one institution varies greatly from another.
By utilizing benchmarks as general guidelines, along with important institution-specific LMS data and custom predictive models, institutions can target and personalize their messages to specific students. Seemingly small nudges, given in-term, are ensuring students stay on their most successful path to graduation.
RELATED: COMMUNITY INSIGHTS ISSUE 1
We collaborated with more than 50 institutions to examine historical records and found that detection rates with triggers, such as cumulative GPA, catch — on average — only 21% of non-persisting students. Meanwhile, our predictive models detect — on average — 82% of non-persisting students on the first day of the term.