How do we use data to measure, observe, and plan better?
Data can come in many shapes and sizes—in tables, from interviews, as GPS (global positioning system) points, via photographs—but they are all observations of the world around us. Through data science and analysis, we aim to understand the patterns, relationships, and factors that influence our city, and use that information to shape evidence-based decisions.
Data science and analysis can help us answer questions like:
- Who is being served by a program?
- Which factor(s) or characteristic(s) most explain differences in an intended outcome?
- Which areas are most likely to need preventative services?
What data science approaches do we have to work with?
Administrative data analysis: Administrative data—data that are collected while a program or service is active, or being administered—are among the most detailed sources of information that we can use. We work closely with our agency partners to identify what data they are already collecting in their work. We can then analyze that data to develop an accurate picture of who is being served by a program or policy, or where a program is having its greatest impact. When we analyze this type of data in a descriptive way, without predictive modeling or causal evaluation, we call that doing “administrative data analysis.”
Predictive modeling: Where we don’t have data—because they are difficult or costly to capture—we can sometimes use predictive modeling to fill those gaps. Predictive modeling uses information that we have to make educated guesses about information that we do not have. These guesses can help us judge how likely something is to happen. This can be hugely helpful in making decisions or allocating resources.
To fuel our analyses, we rely on data from many sources—both inside the government and from external partners. This includes large amounts of government data, which can range from information in paper archives to those stored in modern case management systems. The Lab’s Data Science team brings it all together using data pipelines to inform action and decision making. Whether we’re using data to build maps, dashboards, and visualizations, or using it to write reports based on our findings, everything we do with data is in the service of helping our partners ground their work in evidence.