Skip to main content

About Open Source Analytics

When a data project’s intermediate steps, as well as its end results, are shared, the process can be applied elsewhere. We practice and encourage others to practice Open Source Analytics, meaning that we share the source code for data analytics projects the same way we publish data publicly as Open Data. Sharing the blueprints for data-driven innovation in New York City holds us to a high ethical standard while empowering the community to build on our efforts.

Simply handing over a finished set of metrics at the end of an analytics project limits civil servants’ insight into how the analysis was done and stifles further improvements on the models. The individual steps taken – including what type of analysis was selected and why, insights gained from subject matter experts, and disclaimers on source data – are as important as the results. Data analysts working in government should be held to a high standard of transparency about goals, process, and results. “Open sourcing” analytics efforts not only increases visibility into the way City agencies use data and develop algorithms, it also helps spread awareness of the power of Open Data and data science in municipal service delivery.

Through our Open Source Analytics efforts, we strive toward:

  • Transparency and Accountability: We believe an open source model is more effective than a specific regulatory regime in enforcing the ethical development of algorithms in the public sector.
  • Enabling Replication: Publishing source code allows agency clients to follow our steps to verify our conclusions. It also allows other agencies or jurisdictions to configure our work for their own context.
  • Continuous Improvement: Open-sourcing analytics projects allows MODA’s data scientists to collaborate directly with analysts in adjacent roles across City government, who can improve on the models MODA develops.
  • Better Knowledge Management: The best way to maintain and grow institutional knowledge is by making it discoverable through a Google search. Proactively publishing enforces both quality checks and makes information discoverable across the City.

Our Process