As we discussed in the March 13, 2020, blog: “Modernizing Disclosure Avoidance: What We’ve Learned and Where We Are Now,” we have been working with the data user community on a set of metrics that will allow you to evaluate improvements in the iterative design of the 2020 Disclosure Avoidance System (DAS).

A draft version of the proposed quality measures is now available. This Data Metrics Overview describes each measure and how they work. As a baseline, we applied these measures to the version of the 2020 Disclosure Avoidance System (DAS) used for the 2010 Demonstration Data Products released in October 2019. Those results are in this Data Metrics Tables.

Once we have an agreed-upon set of measures, we will calculate, update and release the new results after completion of each DAS development run. You will be able to compare the latest results against the baseline on an ongoing basis. This will allow you to gauge iterative improvements. It will also allow us to meet the quality standards you expect from Census Bureau data products, while still abiding by the pledge of confidentiality that underlies the 2020 Census.

We welcome feedback and questions on these measures. We’ve recently shared these draft metrics at the first meeting of the small working group of expert data users convened by the Committee on National Statistics of the National Academy of Sciences (CNSTAT). We’ll be engaging with this group to document improvements and overcome remaining challenges in the 2020 DAS. You can access the presentations from that meeting below. Our goal is to release the first update to the metrics on the DAS iterative developments in May 2020.  We are currently evaluating the feedback on these metrics received through the April 24^th comment period. We will still consider comments and suggestions as they are received. Please submit feedback to: 2020DAS@census.gov.