How I Did the Collaboration: Tips for Finding Missing Students
Sharon Lurye reveals how she and colleague Bianca Vázquez Toness collaborated with Stanford University data experts to investigate troubling school enrollment declines resulting in at least 230K missing students during the pandemic.
Photo credits: Allison Shelley for EDUimages (principal visits student home); James Minichello of AASA for EWA
One of the biggest ways that the COVID-19 pandemic affected the education system is that many families ended up leaving it. Public schools in the U.S. lost over 1.2 million students from the fall of 2019 to the fall of 2021 – 2.5% of the entire K-12 population. Many news outlets reported on this massive enrollment drop, but one key question remained frustratingly unanswered: Where did all those children who left public school go instead?
Our investigation revealed that at least 230,000 students had gone missing from both public and private education. That number includes over 151,000 students from California alone, and 59,000 from New York. To find those missing students, we embarked on an unprecedented, nationwide data collection effort, in close collaboration with Stanford University.
Here’s how my team did the story on missing students, with tips on how you can manage collaborations and shine a light on students who have slipped through the cracks in your state.
How the Reporter-Researcher Collaboration Began
The seed of this series started, appropriately, at the Education Writers Association’s National Seminar in Orlando in summer 2022. One of the speakers, Professor Thomas Dee of Stanford’s Graduate School of Education, described his research on enrollment losses in California. We thought: Why not expand that analysis to the entire country?
My team wanted to know: Did the kids who left public school simply go to private school or homeschool instead? Perhaps they moved away, or there were fewer kids to begin with due to declining birth rates?
To find answers, we needed to see how enrollment at nonpublic schools and the school-aged population changed before and after the pandemic. We collaborated with Dee and Big Local News, an organization at Stanford that supports data journalism, to help with gathering and analyzing data.
I am the data journalist on AP’s education team, so I worked on collecting private and homeschool data. Meanwhile, Dee analyzed demographic data from the U.S. Census, and Justin Mayo from Big Local News collected the data on public school enrollment, mostly from federal sources.
There is no centralized source of data on private and home schools in each state, so I reached out to all 50 states individually, then cleaned each state’s data so that it was in a format comparable across states. I cleaned the data using the R programming language, which is a skill I learned at EWA’s Diving Into Data workshop four years ago. Only 21 states, plus the District of Columbia, had data available on both private and homeschool enrollment, illustrating how difficult it is to meaningfully track down students who have left the public school system.
In theory, if enrollment went down in public schools, there should have either been a corresponding increase in nonpublic schools, or a drop in the school-age population. Of course, reality was different from theory. In the states with complete data available, public school enrollment fell by about 700,000 students. At the same time, private and homeschool enrollment grew by 287,000 kids. Population changes accounted for another 183,000 kids.
That still left a huge gap of 230,000 students. There was no data that could explain where they had gone after disappearing from public school – they were missing. There was a high likelihood that they were either in unregistered homeschools or not getting an education at all.
Finding Missing Children
Now we had the data, but data is nothing without the human story behind it. My colleague Bianca Vázquez Toness spent half a year searching for some of those missing students. It’s crucial to find these students, she noted, to prove they exist and that the problem is real.
“It’s important to show that the problem has not gone away for a lot of people. There are people who are still struggling to attend school,” Bianca said.
In addition, telling these students’ stories helps readers understand the wide variety of reasons that young people may leave the school system. Housing instability, mental health troubles, fear of COVID-19, pressure to work instead of going to school, frustration over online learning, and a sense of alienation from the school system were all factors that pulled students away from the classroom.
“Until we actually talk to them and hear from them, we won’t understand what could be pushing all these kids from signing up for school in the first place,” Bianca added.
Bianca’s advice for reporters is to look for missing students through organizations that help vulnerable people.
“The kids who are missing from school tend to be the most vulnerable among us,” she said, so don’t limit yourself to groups specifically focused on education – “I would look everywhere.”
Food banks, service worker unions, immigrant rights groups, and education advocacy groups can help connect reporters to families.
This was the first time I embarked on a project with so many levels of collaboration. I worked with a fellow reporter, with multiple editors, with a researcher, with the folks at Big Local News, and with our graphics/interactives team for the visual presentation of the story.
Collaborations are great because you can divide the workload and gain new perspectives and insight from specialists outside your organization. For example, Professor Dee helped us make sure our methodology was sound, brought a wealth of knowledge and expertise, and made himself available as an expert source for the reporters who wrote localized versions of the story.
However, collaborations also bring their own challenges. When there’s more cooks in the kitchen, you need to put more work into communication and organization to ensure everyone can work together smoothly. One issue, for example, is that we realized that we were using multiple different platforms to share data (Google docs, Github, etc.), which made it more difficult to ensure we were all looking at the same information.
My tips for collaborations are:
Clearly define what everyone’s role and purpose in the collaboration is.
Plan how you’re going to share documents and data, ideally sticking to one common platform.
Be sure to check each other’s work. For example, if there’s data that exists at the federal and state level, check to see if those numbers match each other.
Start bringing people together as early in the process as you can, such as by having a meeting near the outset of the project with all the reporters and researchers.
Throughout this process, we’ve been committed to sharing our data so that other reporters can localize the story for their own region. All our data and methodology can be foundhere. You can use this data to investigate missing students in your own area, diving into questions such as:
What happened to students who left public school in your state or city?
What are the factors that cause disengagement from the public school system?
Who, if anyone, is tracking students who left public school?
Are there engagement efforts and outreach to families to get them to come back to school?
The mission of AP’s Education Reporting Network is not just to publish our own stories, but to help other reporters build on our reporting and dive deeper into the story in their own state or city. Children are still missing, and investigating where they went will be crucial to telling the larger story of how the pandemic has impacted the most vulnerable young people among us.
Sharon Lurye is a data reporter for the Education Reporting Network at The Associated Press.