Formative Assessments With Learning Analytic Solutions
Blue Canary’s Lighthouse Retention Analytics Solution
Blue Canary’s Lighthouse retention analytics solution uses your data to help improve student success. The steps to a successful implementation include:
- Aggregating your historical data to build a custom predictive model.The data would likely come from the LMS and SIS applications, in addition to eBook or other valuable sources
- Creating a model to predict answers to such questions as “will the student pass this class” or “will the student remain engaged this week”.
- Create reports and dashboards for faculty, students, and advisers
- Integrate the dashboards with your existing systems (LMS, CRM). There is no new application to log in to we don’t want to create any roadblocks that might reduce the likelihood that someone will act to help a student.
How Do Student Retention Systems Work?
Retention is an issue at many institutions getting students to enrol is one step, keeping them on track to graduation is another. Concurrently, your institution likely has significant amounts of data that describe student activity. These data can be used to help improve student success. Blue Canary works with colleges and universities to gather, analyse, and distribute data so that students can be helped before it is too late. The Blue Canary Lighthouse solution can answer questions such as “will the student pass this class?” or “will the student attend class this week?”. Lighthouse will integrate with the institution’s LMS or CRM systems to put the answers to these questions in front of those who will act on it and help students.
Requirements
Some of the keys to a successful implementation include:
- An activity data footprint of the student: this includes high level data like demographics as well as more granular data like LMS logins, content access, and the grade-book.
- A robust data history: anywhere from six months to two years depending on the school and its size.
- The desire to act: a predictive model won’t solve all of your problems, no matter how accurate. The institution needs to be willing to take action on the data in order to improve student success
Value
A retention analytics solution pays for itself based on students who remain enrolled with your institution. Implementations and studies have shown that a well-implemented retention analytics solution can increase retention rates by one to five percentage points (depending on what retention activities are currently utilized).
Philosophy
the key to success in leveraging data and analytics is the commitment to a culture of data-driven change. The data won’t tell you WHY a student is at-risk, nor will it tell you who to fix the issue. It’s a timely signal indicating the students who might need help. The institution needs to know who will act on the signal, how they will intervene, and has to be committed to making sure that faculty members and advisers will take the steps needed to help students.
Customer Success
Blue Canary Analytics
Blue Canary Analytics Retention is one area of the higher education life-cycle that stands to benefit the most from analytics. Data collected by an institution has the potential to reveal reams of valuable information about a student – the key is to be able to extract those data and utilize them appropriately. Between 2011 and 2013, a private open admissions university undertook just such an analytics project. The goal was to use student data in order to generate advanced signs of attrition. These data were used by counsellors who communicated with students in an effort to solve the issues and retain the students. A 6-month pilot ran from October 2012 to March 2013 with 15 counsellors and 4,500 students. The pilot showed that counsellors who used these advanced signals had a retention rate that was statistically higher than counsellors who didn’t have access to the signals. The project was rolled out to all 60,000 Associates degree students with subsequent planning for a campus-wide roll-out.
Framing the Problem
Arguably, the most vital part of an analytics project is understanding the specific problem one is trying to solve. A simple statement like ‘we are trying to improve retention’ is a good start, but it’s not specific enough to ensure success.
The next evolution of this thought process is to ask questions like ‘how do we measure retention’ and ‘how might we intervene with students in order to improve retention’? This second question is vital. The key is not to get hung up on doing the analysis. Assume there is perfect data and one could predict student risk. What would the intervention be? How would the institution act on it?
This thought process led the institution to look at attendance. As an open-admit institution with shorter duration courses, weekly attendance is important. Missing a week of attendance is a signal that the student might be encountering issues. In essence, missing attendance is a proxy for attrition (and therefore retention). The analytics project was therefore framed as:
- Predict if a student will miss attendance next week
- If the risk is high enough, forward data to a counsellor who can intervene
Initiating the Project
A common question about analytics projects revolves around getting backing from the organization to initiate. In the case of this retention project, it took over 12 months before the project became ‘official’. The year leading up to the official launch consisted of two work streams. One was building the predictive model and the other was internal selling of the benefits.
Key Takeaways:
- Created multiple versions of the predictive model
- Drafted project team members with domain expertise
- Sold the project from the top (to executives responsible for retention)