Moving data between geography types (points, lines, and polygons) or between different sized geographies (small to larger polygons) is a challenge for anyone who uses data. Here at mySidewalk, we mainly employ two techniques to move spatial data from one geography to another. 


The first and the simplest technique is aggregation, which is simply a grouping of values from smaller geographies into a larger geography. Put another way, the smaller areas are combined (added, averaged, or median taken) to create new values for the larger area. This technique works well for point data, as the point data can simply be aggregated to polygons that contain the points. Another example would be census geographies that nest neatly inside of each other. 

The graphic above shows how U.S. Census geographies nest together. Reading upper left to upper right, then lower right to lower left, the graphic moves from a county to more granular geographies. Of particular interest is how blocks make up a block group and multiple block groups make up a census tract.  

Weighted Block Apportionment

When using non-point data or data that does not have nesting polygons, a more advanced technique of apportionment is used. Here at mySidewalk we use weighted block point apportionment. Apportionment allows you to add the data within mySidewalk to your custom geography.  

A block is the smallest unit of geography used by the U.S. Census. A block point centroid is a computer-generated point that falls in the center of a block. A block centroid contains the ratio of population, housing units, and households of its parent geography, a census block group. These ratios are then used to estimate values for custom geographies. If a block centroid falls within a custom polygon, the population, housing unit, and household ratios are used to estimate the values for data that is available at the block group level. Even if the map view shows your custom polygon overlapping smaller geographies, such as census tracts or block groups, the data is only apportioned if the custom polygon contains the block points. 

Weighted block point apportionment is noticeably different than spatial overlay. When using a spatial overlay, the percent of overlap between the base data and your custom geography is what determines how much data is added to your custom geography. For example, if 60% of a block group falls within your custom geography, than 60% of the values for that block group will be added to your custom geography. 

mySidewalk uses weighted block point apportionment instead of spatial overlay for several reasons. Spatial overlay ignores the distribution of people, households, and housing units, assuming uniform distribution of the values within a block group. In addition, spatial overlay cannot be used for data that are averages, means, medians, or scores.

Helpful Tidbits

The U.S. Census produces the block-to-block group ratio table with every decennial census. The current block-to-block group ratio table used in mySidewalk comes from the 2010 decennial census. 

To receive more accurate results using apportionment, we at mySidewalk recommend including at least 40 block points within a custom geography. A warning message will appear if fewer than 40 block points are used. 

Still unsure how aggregation or apportionment work? Please send us a message using the bottom right hand corner chat feature—we’re here to help!  

Did this answer your question?