In recent years, US school budgets have been relentlessly cut. Teachers have turned to the website DonorsChoose.org to seek private donations for their classrooms. In order to maximize giving, DonorsChoose should make donating as easy as possible. Yet with the myriad of projects seeking support on DonorsChoose.org, it's difficult to decide where to place a donation. Even after narrowing with several filters, decision fatigue sets in when searches result in multiple pages of options.
We propose a simple solution to this hurdle: display an "Impact Score" at each step. By using a statistical approach, Bayes Impact and DonorsChoose.org can reduce information overload, allowing donors to filter by the metrics they care about and quickly determine where their donations will do the most good. The "Impact Score" measures the variables that most correlate with funded projects, meaning that donors won't waste time backing projects that won't ever meet their minimum funding goal.
Our prototype not only provides a solution for donors, but for teachers as well:
For donors, a map-based search makes it easy to drill into a geographic region of interest. The map displays schools as bubbles, sized by Impact Score and colored by common search filters.
For teachers, we provide a location-based predictive model that calculates the impact factors that are most important to donors in their area, enabling teachers to better understand how to tailor their requests to get funded.