Our 311 User Manual Approach

Manual Framework

Our group started this project with the intent of building a new tool that would make it easier to manipulate 311 Open Data so it's easier support other research questions. While investigating how to make the tool, a new problem became clear: most research questions aren't around manipulating the data, but interpreting it. 

User interviews and a market scan helped us develop a framework for developing content. For the manual, we've organized content into areas that address two key areas: 

  1. Analysis: Manipulating the data
  2. Understanding: Interpreting the data 

Analysis: The Analysis portion of our manual serves as an in-depth data dictionary that outlines the new ways in which we've organized the data for easy filtering. Data is re-organized in spatial, temporal, and categorical considerations form our research.

311 Open Data is has dozens of columns, but not all are useful for each research question. Some of the simplest filters may be derived from other columns, but are difficult to create due to the shear size of the dataset. We used SQL to essentially add columns to the dataset to make filtering for key information simple. Our additional columns answer questions we heard in our research:

  • Weekend/Weekday filter: "I'd like to quickly compare noise complaints on weekends and weekdays"   
  • Police Precinct filter: "I'd like to merge noise complaints with crime data"
  • Master Categories: "Show me all the complaints that have to deal with the heating and the home" 

Understanding: The Understanding portion of the manual focuses on providing context regarding how complaints are resolved, categorized, and geocoded. It also aggregates resources to provide inspiration on how the data can be used. Content in this section helps to answer questions like: 

  • What happens with geographies change over time? 
  • How do new categories emerge? 
  • When something is "resolved," what does that mean in practice? 

The answer to most of the above is that the data is often wildly inconsistent. The data is still incredibly valuable, but raising considerations will help researchers be able to apply it in a more meaningful way. 





User Research Framework

Pull Requests / Feedback Instructions