Big data has created complex new challenges to data privacy. One advantage of administrative big data is the enhanced feasibility of large scale record linkage. How can we make more data available to inform decision making without creating “Big Brother”? How can we inform this needed revolution in privacy protection without cutting back access to data?
In this webinar, Cavan Capps and Micah Altman will review their comprehensive analysis of an ACS use case that can be used to inform key decisions on how to protect data privacy while leveraging the latest data technologies. The results suggest that a multi-tiered access system to the data may be warranted in the future, potentially including traditional tabulations and regressions protected by Differential Privacy or variants of Secure Multi-party Computing (SMC) in software or in hardware using SGX, among other options. The webinar will discuss some of the strengths and weaknesses of the tools mentioned above and propose how such an infrastructure might be constructed.
Finally, the webinar will provide an update on our work continuing work to examine the practical use of SGX-SMC and software based SMC for data collection and integrating shared confidential data from different sources. This enables data sharing while maintaining individual privacy of individual during any analysis. Differential privacy will be used to ensure that any outputs remain confidential.
Micah Altman, Head Research Scientist, MIT Libraries
Cavan Capps, Big Data Lead, U.S. Census Bureau
The Census Bureau is introducing a new framework to protect individual data in the Decennial Census: “Differential Privacy”. This has implications for the reliability and availability of invaluable federal statistics – decreasing accuracy for small areas and small sub-population counts and reducing the scope of various data products in exchange for improved privacy protections.
This webinar from the Association of Public Data Users will provide a background on Disclosure Avoidance, details on the policy decisions leading to Differential Privacy and its subsequent implementation, and comparisons of recently released data comparing previously available 2010 Census data with data demonstrating the impact of Differential Privacy. Register today to learn more.
Beth Jarosz, Senior Research Associate, Population Reference Bureau
Kathryn Pettit, Principal Research Associate, Urban Institute
David Van Riper, Director of Spatial Analysis, IPUMS
Five years ago, staff at the Bureau of Labor Statistics had to read and manually code hundreds of thousands of written descriptions of work-related injuries and illnesses each year. Today, more than two thirds of these codes are now assigned by a deep neural network, which evaluations suggest is substantially more accurate on average than trained human workers.
In this webinar, Alex Measure will discuss how BLS addressed some of the many challenges inherent in this transition. Attendees will learn:
how BLS built these new computer systems
how they decided when and how to use them
how to evaluate their performance
how BLS monitors and maintain them to continually improve performance.
Alex Measure, Economist, Bureau of Labor Statistics
When the U.S. Commission on Evidence-Based Policymaking issued its unanimous recommendations to Congress in 2017, it called for the exploration of new approaches that promote data access and privacy preservation at the same time. This webinar discusses an application of one such technology – multi-party computation – in a real-world setting to assess the applicability of the approach in public agencies.
A demonstration project in Allegheny County, Pennsylvania applied privacy-preserving approaches to generate responses to policy-relevant questions about mental health services, homelessness services, and other public health policies. This demonstration project offers a compelling example of how the technologies can be deployed—which can advance consideration of the approach within agencies at all levels of government. Register today to learn how this new technology could impact the data you rely on.
Presenter: Nick Hart, Ph.D., CEO, Data Coalition Fellow, Bipartisan Policy Center
USDA Research Relocation and Reorganization: Perspectives from Former USDA Chief Scientists and Administrators
Agriculture Secretary Sonny Perdue in August caught Congress and USDA stakeholders by surprise with a proposal to relocate the USDA Economic Research Service (ERS) and the USDA National Institute of Food and Agriculture (NIFA) outside of the Washington, DC area and to move ERS from the USDA research arm to a policy arm.
The proposal has raised many questions: Will there be a chance for the USDA partner community to comment? Was the partner and stakeholder community previously consulted? Why the speed of action? What is Congress’s role? What is the problem being addressed? Many concerns have also been raised: Will NIFA and ERS relevance and reach be impacted by being moved away from key audiences and policymakers? Will ERS’s standing as a world premier economics research institution be retained with the expected staff attrition for those not willing to move their families? What will be the impact of ERS’s perception as an independent, trusted source of information and analysis be impacted by its move to a USDA policy office?
The panelists will address the many questions and concerns raised about the USDA proposal; address viewer questions; and suggest actions for Congress, USDA, and viewers.
This June the White House released a proposal that would shift the Bureau of Labor Statistics from the Department of Labor to the Department of Commerce, joining with the Bureau of Economic Analysis and the U.S. Census Bureau under a single economics and statistics agency. Join us for a facilitated conversation with Dr. Ken Poole as he talks with Dr. Nancy Potok, Chief Statistician of the United States at the Office of Management and Budget and Dr. Erica Groshen, former Commissioner of the Bureau of Labor Statistics. Attendees will learn more about the reorganization from Dr. Potok and Dr. Groshen, better understand the rationale for the proposal, and gain insights about key considerations that still need to be addressed. This is your chance to provide feedback about your concerns. You are strongly encouraged to bring questions, concerns, and ideas to the conversation.
Dr. Nancy A. Potok, Chief Statistician of the United States, U.S. Office of Management and Budget
Dr. Erica Groshen, Former Commissioner, Bureau of Labor Statistics
Dr. Kenneth E. Poole, CEO, Center for Regional Economic Competitiveness
The Data Documentation Initiative (DDI) is an international standard for describing survey and other social science data. Documenting research data with DDI facilitates interpretation and understanding — both by humans and computers. Learn how DDI improves the ability of researchers and organizations to Document, Discover, and Interoperate in this APDU Webinar on March 7.
Presenters: Barry Radler, Distinguished Researcher, University of Wisconsin
Institute on Aging Jon Johnson, Data Management and Metadata specialist working, UK Data
Jared Lyle, Archivist, Inter-university Consortium for Political and Social Research (ICPSR)
With the 2020 Census fact approaching, both the Census Bureau and outside stakeholders are working to ensure that the Census is fair and accurate. In turn, both groups are developing mapping tools to identify areas that are difficult to count.
The low response score (LRS) is a metric developed by the US Census Bureau to classify geographic areas according to their propensity to self-respond to surveys and Censuses. Nancy Bates of the Census Bureau will showcase a new publicly-available tool that greatly simplifies the use of the LRS using a web-based mapping platform. The platform known as the Response Outreach Area Mapper, or ROAM, allows users to select a geography of interest to display tract-based maps colored-coded according to LRS. In addition to quickly identifying hard-to-survey areas, the ROAM also informs users why a particular area may be hard-to-count.
The Center for Urban Research at the CUNY Graduate Center, working with the Leadership Conference on Civil and Human Rights and other census supporters, developed the “Census 2020 Hard to Count Map” at www.CensusHardtoCountMaps2020.us to provide community groups, the media, Census Full Count Committees, and others with an online tool to highlight the hardest to count tracts in the country. This presentation by Steven Romalewski will introduce this intuitive, easy-to-use tool and potential upcoming additions.
Presenters: Nancy Bates, Senior Researcher for Survey Methodology, Research and Methodology Directorate, U.S. Census Bureau
Steven Romalewski, Director, Mapping Service at the Graduate Center / CUNY
18M open datasets exist today, and growth is accelerating. But these data sets live in data portals without common taxonomies or architectures, and must first be cleaned and prepared by data users. Human and computers normalize, extract meaning, and identify correlations, but this work is siloed: used for one project, then lost forever, only to be repeated from scratch by the next person to touch the data.
Open data can help us rise to humanity’s toughest challenges, but only if we maximize its network effect. To build the web of Linked Data, we have to start by connecting the people who are working with data.
Patrick McGarry of data.world will answer the following questions (and more) in this webinar:
How can we enhance open data that has already been published?
How are leading open data publishers combining the power of linked data and open data?
What is the role of public-private partnership today, and how should it evolve?
Presenter: Patrick McGarry, Head of Community, data.world
Many organizations collect data to help manage and monitor the performance of their programs. This administrative data can also be used to improve program evaluation and management and produce original research. As a first step, organizations should ensure the data is of a high enough quality to support research and evaluation. The webinar will serve as a primer on how organizations can improve the quality of administrative data for research. The webinar will cover:
The importance of administrative data quality,
The major issues of data quality,
Strategies for reviewing the quality of the data, and
Strategies for cleaning the data.
The webinar will cover major issues of data quality, including: units of analysis, missing values, invalid values, incorrect formatting, and value inconsistencies both within and across variables. In addition, the webinar will provide specific and concrete strategies for reviewing and cleaning the data in preparation for research. The webinar will be particularly useful for organizations that are relatively new to using administrative data for research and evaluation.
China Layne, Ph.D., Manager, Data Analytics and Research at Summit Consulting