NSF, Census Bureau Announce Research Network
The National Science Foundation (NSF) and U.S. Census Bureau have awarded interdisciplinary research grants aimed at finding new ways to interpret social, behavioral, and economic data and new ways to use and disseminate the resulting statistics.
The following eight research awards were made to establish a set of research nodes that will conduct long-term, interdisciplinary, methodological research and educational activities.
Data Integration, Online Data Collection, and Privacy Protection for the 2020 Census
Principal Investigator: Stephen Fienberg, Carnegie Mellon University
This project will conduct research on three basic issues of interest related to collecting decennial census data: privacy, costs, and response rates. Researchers will address the practical problems of ensuring confidentiality and privacy while producing useful statistics for public and private purposes. They will investigate the use of administrative records to create the basic census frame, as well as other possible uses of administrative records as part of the census. This node also will conduct experiments that implement new ways of encouraging participation in an effort to reduce the decline in (or perhaps even increase) response rates and examine the unit used to collect census information (currently the household).
Improving the Connection Between the Spatial and the Survey Sciences
Principal Investigator: Seth Spielman, University of Colorado
This project will exploit new forms of geographic information and recent advances in spatial statistics, a type of statistics that focuses on the relationships between variables for areas in close proximity. In this case, the project is intended to make small-area estimates from the American Community Survey more accurate. The project also will foster an improved connection between the spatial and survey sciences, yielding both immediate and long-term benefits for the estimation, dissemination, and usability of the small-area statistics produced by the U.S. Census Bureau and others.
Integrated Research Support, Training, and Data Documentation
Principal Investigator: John Abowd, Cornell University
This project will develop a curation system designed and implemented in a manner that permits synchronization between the public and confidential metadata about the data sets available to researchers from the U.S. Census Bureau. The scholarly community will use the system as it would a conventional metadata repository, deprived only of the values of certain confidential information, but not their metadata. The researchers also will improve imputation methodology for the Longitudinal Employer-Household Dynamics Program.
Enhancing Federal Agencies’ Data Dissemination Capabilities
Principal Investigator: Jerome Reiter, Duke University
This project will develop broadly applicable methodologies that will transform and improve how statistical information is shared with the public by the federal statistical system. In particular, researchers will advance methodologies and tools for disseminating public use data with high quality and low risk of confidentiality breaches by developing theory and methodology for releasing multiply imputed “synthetic” data sets based on flexible, nonparametric Bayesian models built specifically for complex data sets.
Linking Surveys to the World—Administrative Data, the Web, and Beyond
Principal Investigator: Matthew Shapiro, University of Michigan
This project aims to improve survey measurements of economic and demographic data and potentially supplement or replace surveys with statistics based on administrative, web-based, and geospatial data. Applications include using linked survey-administrative data to assess attrition, selective nonresponse, and measurement error in surveys; using web-based social media to measure job loss, job creation, small business creation, and informal economic activity; using administrative geospatial data to enhance small-area estimates; and investigating the relationship between public use, synthetic, and internal versions of the same data sets.
Development of Innovative and Transformative Approaches to Data Collection
Principal Investigator: Allan McCutcheon, University of Nebraska
This project will focus on improving survey data collected from computer-assisted methods. Objectives include evaluating the use of four diagnostic tools for identifying measurement errors in computer-assisted, interviewer-administered data-collection instruments; evaluating the use of adaptive/responsive designs in which a dynamic modeling of collected data is used to modify the questionnaire as the data are being collected; and evaluating the application of calendar- and time diary-based data-collection methods to aid in the accuracy of behavioral self-reports by tailoring questions to the needs of individual respondents.
Census Bureau Data Programs as Statistical Decision Problems
Principal Investigator: Bruce Spencer, Northwestern University
This project will address fundamental problems for all government statistical agencies, such as how to understand the value of the statistics they produce, how to compare value to cost in order to guide rational setting of statistical priorities, and how to increase value for given cost. Researchers will extend and apply statistical decision theory, including cost-benefit analysis, to attack such basic questions.
In addition to generating new research methods and using advanced research practices and procedures, the researchers will address social and economic issues addressed by the federal statistical system. These issues include improving survey data collected from computer-assisted methods, exploiting new forms of geographic information, and ensuring confidentiality of the data collected and reported.
One more important result may be improved cost-efficiency and quality for the data collected in surveys. Improving censuses and surveys for the Census Bureau and other statistical agencies results in more accurate statistics for policymakers and officials to draw more precise conclusions and make better-informed decisions.
“This grant program gives the Census Bureau and the entire federal statistical system the opportunity to leverage the expertise of academia to solve problems we face every day in delivering cost-efficient statistics and information to the public,” said Robert Groves, director of the Census Bureau. “This research is an investment that will lead to cost savings, and we are excited about the possibilities for learning from our colleagues and for collaboration over the next five years.”
The projects also will foster the development of the next generation of researchers with skills relevant for the measurement of economic units, households, and people.
“These awards provide a unique opportunity for researchers to advance fundamental understanding of important issues related to the collection, analysis, and dissemination of data in the social, behavioral, and economic sciences within the context of salient problems for the federal statistical system,” said Cheryl Eavey, program director for NSF’s Methodology, Measurement, and Statistics Program.
More information can be found at the Census Bureau’s website.
Improving the Interpretability and Usability of the American Community Survey Through Hierarchical Multiscale Spatio-Temporal Statistical Models
Principal Investigator: Scott Holan, University of Missouri
The American Community Survey is an ongoing survey that releases statistics and estimates annually, providing communities with the timely information needed to plan the distribution of resources and services. This project will improve the interpretability and usability of the survey estimates, in particular the estimates for small areas and small population groups, through the development of statistical models that take account of both changes over time and differences over geographical space. In addition, researchers will provide a variety of methods that are of independent interest and can be used in many other surveys administered by the Census Bureau and other federal statistics agencies.