The first test of four tests planned for 2015 was the Address Validation Test (AVT) which was in the field from September 2014 through January 2015.  This test enabled us to assess the performance of the methods and models that will help us develop the 2020 Census address list and define the address canvassing workloads needed for the operational design decision point in September 2015.

The test contained two components. The first, a full-block canvass that assessed the ability of two sets of statistical models to predict blocks that have experienced address changes that have not already been captured in the Master Address File (MAF). To do this, we collected address data in a national-level address canvassing operation covering 10,100 blocks.

The second component of the AVT was to assess a methodology that uses geographic specialists (a) to detect unrecorded changes in blocks and (b) to identify portions of blocks where change is likely, thus enabling a partial-block canvass that could reduce workloads and cost. To detect possible changes within blocks, staff conducted an interactive review of various materials–primarily aerial imagery over time and geographic quality indicators.  For a sample of those blocks (limited to 600-1,000 blocks for this test), field staff then updated the address list and maps in the portions of those blocks where we identified possible changes.