Per-pupil expenditure is a new requirement that schools must report starting the 2018-2019, as part of the Every Student Succeeds Act. The goal is to provide transparent financial information—for parents, community members, administrators, policymakers, and researchers—to ensure that funding is equitable and effective.
Read more about the challenge here.
As part of the challenge, the Department of Education and Data Quality Campaign provided three scenarios, which we leveraged for personas and user stories.
- Julia saw an article in the newspaper that described how per-student spending in her child's schol is half as much as a school in the next neighborhood. She wants to better understand why that might be the case.
- District Leaders in District A are considering some changes in how they allocate funding between schools. They are interested in athering information on how District B allocates its funds so they can understand similarities and differences between their spending decisions.
- Dr. Hsu, a researcher from a local university, is making a presentation to the local school board on effective uses of federal dollars and wants to share a visualization of expenditure data in relation to school outcome data. Dr Hsu also wants to make sure that the appropriate context and caveats are included in the presentation.
For the purposes of the 2-day challenge, we decided to focus on the first 2 senarios with Julia (a mom) and district leaders.
First iteration — Initial data visualization explorations
Using Tableau, I prototyped many visualizations to explore which ones were most legible and useful.
- Because there can be many external factors to funding & spending that is not represented in the data (age of building, economies of scale), we decided to normalize the visualization by percentage of overall spending, rather than showing just the absolute numbers.
- Some categories are fixed costs, others scale with the size of the student population. We color-coded fixed and variable spending.
Feedback and insights from subject matter experts
Attending the event were also subject matter experts. Based on their feedback, we prioritized information and refined the UI.
Policy maker feedback
- Disaggregated data is a minefield for unintended consequences, including discrimination, categorizing, and stereotyping. A user needs to be able to drill down into the categories, but at the highest level, spending can be categorized into two groups: Learning (teacher salaries) and Operations.
- A school's progress over the last 3 years is more useful than comparison to the school/state.
- At most, a superintendent is comparing to other nearby districts, not to the entire state.
- Parent care most about what the data means for their own individual child.
- "Above" or "below" language is harmful, and not a useful comparison. Average is not a useful benchmark because half the population will always be below average.
Second iteration — Designing the interface
The interactive visualization uses the overview, zoom & filter, details-on-demand design pattern. In the overview, users can compare states, districts, or schools. They can then zoom in on a specific school to view their report card.
On the school report card, one design principle that we wanted to keep in mind is to tell a story. Taking inspiration from data journalism, the site is structured is a question & answer format, prioritizing the perspective of the parent.
One data point alone doesn't tell a full story, so it was important to add context. When comparing against other schools, we offer at least one additional measure of comparison, such as school enrollment size. Finally, in order to prevent parents from making interpretations with incomplete data, we included a letter grade for school spending at the top, which conceals a more complex algorithm for benchmarking performance.
Reflections on working with policymakers in the design process
- Regulations tend to state the minimum requirements. There are many ways to innovate beyond what's dictated by regulation.
- A federal policy can be implemented in various ways at the state and local level, depending on the different needs and agendas of the local government and community.
- At the local level, there are many more nuances that impact student expenditure that must be considered. For example, in Texas, some school districts pay into society security, which results in higher payroll expenses.
- People working in education are there for the same reason—to help the kids. There tends to be more moments of agreement than disagreement.