Let's say you're a college - and you want to know what factors influence graduation rates. You have a Quickbase table with 15,000 students - 6,000 of which have disenrolled, 4,000 that graduated, and 5,000 that are currently enrolled.
Your question (obviously) is "why do students dis-enroll?" - and what you can do to help the 5,000 currently enrolled students graduate.
You believe the answer may lie in the 20 data points you've collected over the years on each student. Some of these data points include:
The "solve for" here (outcome) is a checkbox field in the Quickbase student record called "graduated" (indicating that student has graduated).
So, we want to analyze all the student records with graduated checked, to see if there are common positive (and negative) factors across those 4,000 successful students, and to rank those factors in importance.
Quickbase AI can help. This AI implementation allows you to specify what records and fields you wish to have analyzed, and turn Quickbase loose to do the analysis.
What you end up with begins with a list of factors the AI thinks are important to achieving the graduated outcome. In this case, those factors looked like this:
What you see above is what the AI determined to be the most significant commonalities across all the graduated students' records.
There's more detail here for the outcome "graduated":
Where everything comes together is the way we applied the AI's analysis (real-time) to every currently enrolled student's record.
In this case, we programmatically assigned each enrolled student a "Graduation Prediction" (a percentage predicting graduation probability) - and also included a visual indicator image showing red, yellow, and green based on their risk of disenrolling. This analysis is performed real-time on every record as it's added or edited in the Quickbase table.
Finally, we assigned each student to a "Risk Group".
This allowed us to create a custom-tailored "Graduation Assistance" package for each student. We used Quickbase formulas (or more AI) based on each student's unique situation to help the school design a unique program for each high-risk student - one tailored to help to improve their specific situation, and the odds that they'll graduate.