Project description

We wanted to look into teacher diversity in the state of Washington and how greater representation can improve the outcomes for students of color.

First we needed to show the current state of demographics for teachers and students – and highlighted how few teachers of color there were compared to students of color. We then created an index to score teacher-student representation and showed that by all students of color and by individual races on an interactive map for every school district in the state. Finally we explained why the absence of representation matters and how it effects high-school completion, discipline rates and test scores for students of color. Our audience was the general public in Washington state, especially educators and parents.

Though teacher diversity is an issue that has been covered widely at the national level, it had never been closely examined in Washington. We reviewed tens of thousands of data points, spanning six years and more than 300 school districts and educational facilities.

To produce our findings, we had to create a database linking hundreds of data sets containing teacher and student demographic information. No one in the state — not even in the Office of Superintendent of Public Instruction — had done such a thorough analysis, at least not one that had been made public.

What makes this project innovative?

The three-part project was a collaboration between two regional newspapers (The Seattle Times and The Columbian), resulting in a more robust reporting that better serves Washington residents. This project was also really driven by the data. We created a teacher-student representation index and calculated a score for every school district in the state. No other index has been created at the statewide scale for this topic. And we went even further to score each district’s representation by individual races and by all students of color. Our data and graphs were the foundation of our project and we used them to show our readers the demographic mismatch instead of just telling them about it.

What was the impact of your project? How did you measure it?

Our goal was to illustrate the lack of diversity in the teaching profession in an accessible way, and create a resource that’s easy and quick to reference using the many charts in the stories. As a result, public officials at the local and statewide level have referenced the series in their proposals for more funding to recruit teachers of color. Nate Gibbs-Bowling, a finalist for national teacher of the year, called the series some of the best reporting on the topic of teacher diversity in the country. It also won best story of the week in a newsletter penned by education journalism critic Alexander Russo.

Source and methodology

Using Structured Query Language (SQL) to combine spreadsheets, we created a database that linked teacher and student demographics for 295 school districts in the state for the last six years. Most of the reporting and charts in the series came from this database, which spanned several thousand lines. The student data file was readily available on the state education department’s website (Office of Superintendent of Public Instruction), and the teacher data was obtained through public records from the state. We then cleaned the data sets to eliminate districts where no data was available in either category, or where it was apparent the state made an entry error. To check the accuracy of the database, we manually added up the totals instead of relying on the totals provided by the state. We also spot-checked student and teacher counts for each district on the state department website to make sure no errors were inserted when the data was cleaned. For the interactive map, we used a ratio (the percentage of teachers of color divided by the percentage of students of color) to show how closely the teaching staff at each district resembled the student demographics. The closer to 1, the more representative. We used a color scale to show proximity to 1. We also published an article detailing the data methodology -

Technologies Used

The first part was custom built using HTML, CSS, and Javascript. The static graphs and charts were created in Adobe Illustrator with multiple versions to accommodate different screen sizes. The interactive map was a customized Leaflet map with a GeoJson file created in QGIS. SQLite Studio and Excel were also used during the data analysis.

Project members

Dahlia Bazzaz, Katie Gillespie, Emily M. Eng, Jennifer Luxton, Joy Resmovits, Laura Gordon, Jane MacDonald, Mohammed Kloub



Click Follow to keep up with the evolution of this project:
you will receive a notification anytime the project leader updates the project page.