What is Dropout Detective??
Dropout Detective is a student retention and success solution that integrates directly with Canvas to provide a “risk index” of how likely it is that each online student will drop out of or fail their online course(s). The program analyzes past and current behavior to predict future performance. This program pulls together the different reasons a student is at risk and makes it easy to quickly go through and look at what might be happening with a student and determine an appropriate intervention strategy.
Analyze Students' Behavior
Today, many schools are managing student retention by reacting to the students’ behavior. And by the time you identify a serious problem, it can be too late to keep that student in class. You are working by looking in the rear view mirror because you can only see what has already happened. Dropout Detective™ replaces that rear view mirror with a crystal ball. We analyze the information that you already have and use our proprietary technology to identify your students that are at most risk. Dropout Detective™ allows you to pro-actively manage retention by alerting you to behaviors and patterns that lead to high student attrition. With Dropout Detective™, your team can work in a targeted and efficient manner to support the students that need the most help.
How it Works
The past behavior of a student in an online course can tell you whether they are going to succeed. Dropout Detective™ integrates with your existing learning management system (LMS), which already tracks these behaviors. The system retrieves the necessary data, applies the algorithm, and populates an easy-to-use dashboard to the administrators of the institution. Students are given a Risk Index and are displayed in a red-yellow-green “stop light” format to draw attention to the most at-risk members of the student body. Administrators can then drill into the individual student’s profile to determine the risk reasons and to develop a pro-active plan of action. Through search capabilities and filters, the institution can target its efforts by assigning the right staff to the right students.
Why It Works
Dropout Detective™ was developed over a number of years by online education professionals looking for a better, smarter way to focus their scarce resources on their most at-risk students. Starting with a checklist on a sheet of paper, these educators identified the behaviors exhibited by both successful and unsuccessful students. Relative scores were applied to the behaviors and were then tweaked over time to produce the most accurate predictive modeling. This scoring system migrated to a spreadsheet and proved to be an invaluable and accurate tool for administrators. Now, Dropout Detective™ can integrate directly with your LMS, retrieve the relevant data points from your records and apply risk weightings to all of your students. The dashboard will allow you to drill into each student’s profile and determine the critical behaviors that have contributed to a high-risk rating.
Background on our Algorithm
The algorithm behind the Dropout Detective takes into account a number of student behaviors and then applies weights to each of those factors. And depending upon other factors, those weights can change.
Some of the factors that we pull from the LMS include, but are not limited to:
- Timestamp for start date of course
- Timestamp for end date of course
- Timestamp for last time student logged into LMS
- Current grade for student in course
- Final grade for student in course
- Total assignments for student in course
- Total assignments turned in on time for student in course
- Total assignments missing for student in course
- Timestamp for last time student accessed the course
- Timestamp for last time student sent a message in course
- Timestamp for last time instructor sent a message to student in course
- Timestamp for post in discussion forum in LMS
- Timestamp for each submission in LMS
Weights applied to these factors can change based on other indicators such as how far along in the course timeline we are, whether or not there are many assignments or just a small number and other such variables.
