When working with data, there are a lot of choices to be made. Detailed below are the methods the mySidewalk team leverages to get data into custom boundaries and how we make historical data fit into new boundaries.
There are main 2 reasons the numbers you are seeing in mySidewalk may differ from other sources:
- We use weighted block to block group apportionment.
- We harmonize historical data to the most up-to-date geographic boundary.
1. We use weighted block to block group apportionment.
You can upload or draw a custom boundary within the mySidewalk Platform. When you do, the mySidewalk Platform can apportion data automatically into your custom boundary.
This works by leveraging 2 things:
- The Decennial Census block ratio tables, which provide the ratio of population, households, and housing units from each census block to a block group.
- Data available for Census block groups.
When you have a custom boundary, the mySidewalk Platform looks up all of the contained Census blocks. It then uses the population, household, or housing unit ratio to apportion the value from the Census block group into your custom boundary. This method provides a more granular way to get Census block group data values into your custom boundary.
Why use the block to block group ratios?
Where people live is not evenly distributed across a Census block group. There may be parks, industrial districts, or retail areas. The Census block ratios reflect this uneven distribution so that we can more accurately take the values from Census block groups and calculate them for your custom boundary.
Which of the ratios - population, households, or housing units - will be used?
Most Decennial and ACS Census data falls into one of those three categories - population, households, or housing units. A household is a group of people living together; a group of people. Whereas housing units describe the physical structure in which people live. The mySidewalk team maintains a metadata ecosystem that keeps track of the best fit ratio type.
Pro-Tip: The accuracy of a single or even a dozen census blocks can suffer from errors. That’s why when you use mySidewalk, you will see a warning if your geography does not overlap at least 40 Census blocks and at least 3 different Census block groups.
2. We are using the most up to date geography.
Census geographic boundaries update at varying intervals and the team at mySidewalks works to make sure we have the most up-to-date set available.
Part of using the most up-to-date boundary is the harmonization of historical data into these new boundaries. This involves taking historical data which was produced for a different boundary and recalculating the historical data so that it fits in the modern boundary, as shown in the diagram below.