Understanding New Jersey’s schools through data

NJ School Data is a collection of free, interactive tools for making sense of New Jersey’s public schools and districts.

Page reviewed July 2026

The New Jersey Department of Education and other entities publish a tremendous amount of information about our schools and districts, but it is often left tucked away in database files on government websites.

This site attempts to make these datasets more useful to parents, educators and others invested in improving outcomes for our schools and students.

I’m Marc Lavallee, and I make these tools. I’ve spent a long career at the intersection of journalism and technology, building things that help people understand the world through information. I’m also a Maplewood, New Jersey, parent with a personal stake in the success of our state’s students and the schools that nurture them.

I am building this site to sharpen my view of our schools, and I’m sharing my work to help others on similar journeys.

Data is only one part of understanding our education system. It can often show what is happening, but it can rarely explain why. This site provides starting points for conversations; it does not try to answer all the questions that arise.

I am using artificial intelligence tools like Claude Code and OpenAI’s Codex to assist with parsing data, generating visualizations, and building a user-friendly website. When I say “we” anywhere on this site, I mean me working together with these tools.

Here are some things to know about how this site has been built so far:

• All of the data you see comes directly from official sources in database-friendly formats like Microsoft Excel files and CSVs. Large language models are not “generating” any of the data here. Nothing is hallucinated.
• That does not mean this site is bug- or error-free. Source data is prone to errors, especially with the kinds of statewide datasets that are often assembled from the self-reporting of hundreds of individual school districts.
• Most government data also has quirks and idiosyncrasies, especially with datasets maintained over long time periods. Changes in policy and society affect what’s recorded and how, and data labels often have nuanced and complex meanings.
• In some cases, I’ve made judgment calls on how to represent these quirks, leaning on my 25 years of data and journalism experience.
• Where we apply statistical modeling to data, we write explainers. There are many ways to interpret data. Rarely (if ever) is there one perfect approach. The goal of detailed explainers is to showcase the judgment calls and trade-offs.
• In terms of the text you come across on this site:
  • Pages and sections that carry my name (like this one) were written by me. For authored text, I often use AI to copy edit, with instructions to suggest improvements for flow and readability but not change meaning or intent.
  • Templated text (chart descriptions and summaries, guiding text, etc.) is mostly generated by AI and proofed by me.
  • The detailed explainers are mostly written by AI, and they’re mostly written for other people’s AI tools to follow along.

A small example of what that work looks like. For three school years, New Jersey’s enrollment files misspelled a single column heading — “Eighth Grade” written as “Eight Grade” — and the software reading them, looking for the correct spelling, recorded no eighth-graders at all, in every school in the state. The students were never missing; one dropped letter hid them. We found it, traced it across every year, corrected it, and wrote it down in the data explainer. A good deal of this site is built from work like that.

As AI best practices are rapidly evolving, this is all subject to change.

If you have questions or comments, or see anything amiss, please do not hesitate to reach out at [email protected].

See the corrections and data integrity policy for how material errors are handled and the public record of corrections made.