Between 2010 and 2014, there were 22 fatalities in New Orleans caused by structure fires, including a tragic Broadmoor house fire in November 2014 that killed five people, including three children. In nearly all cases, no smoke alarms were present. To help guide the City's efforts in reducing fire fatalities, the Office of Performance and Accountability teamed up with the New Orleans Fire Department (NOFD) to develop a model that identifies the neighborhoods most at risk for fire fatalities to help prioritize NOFD's efforts installing free smoke alarms in homes across New Orleans. The model helps identify neighborhoods with homes that have two key characteristics: least likely to have smoke alarms and most likely to experience fire fatalities. For the first characteristic we analyze the American Housing Survey (AHS), which includes a variety of questions regarding housing and demographic characteristics. One of the questions in the AHS asks if a house has a smoke alarm installed, allowing us to perform modeling based on the data. Unfortunately, the AHS only goes to the parish level, so to use the data for targeting, we also must translate the results of a model to a more discrete geography. To do this, we use the American Community Survey (ACS), which is available at the block group level. To model fire risk, we account for the frequency of fires in different parts of New Orleans as well as research from the National Fire Protection Association (NFPA), which shows that individuals under 5 and over 65 are most at-risk of suffering a fire fatality. Based on these risk factors, we create a combined risk map, and identify the fire zones that are most in need of smoke alarms.
forked from cno-opa/smoke-alarm-outreach
-
Notifications
You must be signed in to change notification settings - Fork 0
roddtalebi/smoke-alarm-outreach
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
Predictive analytics for smoke alarm outreach in New Orleans
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published
Languages
- R 100.0%