The question
Is The Hardy Center's tutoring program working, and not only being used?
The method
I combined attendance records, standardized test scores and IXL platform data from separate sources into one clean dataset. I then compared how much each student took part in tutoring with how that student did in school. The results sit in a dashboard that staff without a technical background can open and explore themselves. Docker keeps each metric calculation in its own container, so the calculations cannot interfere with each other.
Tools and methods
Tools
PythonStreamlitDocker
Methods
Data cleaning and mergingCorrelation analysisInteractive dashboardContainerized metric calculation
Limits
The result shows that more participation goes with better outcomes. It does not prove the tutoring caused the gain, because students who choose to attend may differ from those who do not.