Combining estimates for multiple geographies or for subgroups
A few times people have asked us how to use Census Reporter to "roll up" group of Census tracts into a single set of estimates for the combined areas. The approach for this is the same for all geographies, whether tracts, counties, etc.
For estimates of counts, it starts off pretty easy: you can just add the estimates together. However, if you are taking care to pay attention to the margin of error, computing that is not as simple as adding the error for each estimate together.
Note that the approach for combining estimates for different tracts
is identical to combining different columns for a given table. For
example, table B01001 Sex By Age
separates each age bracket into two estimates, one for male and one for
female. It also has very many age brackets. If you want to combine the
sexes, and/or combine brackets to have fewer, you have to handle
combining the margin of error in the same way as if you combine
estimates for different geographies.
The Census Bureau documents various formulas for aggregating error in chapter 8 of Understanding and Using American Community Survey Data: What All Data Users Need to Know. We created a Google Sheets document that implements the correct method for computing the poverty rate and margin of error for that rage for five tracts.
For people who are looking for more information about finding data for specific Census Tracts, we recommend watching our short video, Using Data Search in Census Reporter.