An Overview of Statistical Methodology at the Census Bureau

Research developments at the
Census Bureau
Roderick J. Little
Associate Director for Research &
Methodology and Chief Scientist
Bureau of the Census
• Director Groves vision for research
• Structure and strategic objectives of new
Research & Methodology directorate
• Outreach to external research community:
– NSF/Census Research network
– Research Data Centers
• Some research priorities
Research at the Census Bureau…
• … More important to us now than ever
• Increased demand for data products,
questions asked are becoming more
complex and hard to measure.
• Surveys and censuses are increasingly
expensive and challenging to mount.
• Research is essential to reconcile these
What is the R&M Directorate?
Strategic objectives 1
Build a Research & Methodology Directorate that
fosters innovation and plays a strategic role in
Bureau activities
Develop existing R & M members and recruit new
talent to be thought leaders in research
• Increase peer reviewed research
– Provide an environment and resources conducive to
high quality research.
– Form cross-cutting teams with other Directorates to
solve key methodological problems
Strategic objectives 2
• Porting research on new products/processes to
program areas - especially more methodologically
complex products/processes.
• Good ideas don’t implement themselves
• Assist with the refinement and implementation of
innovative ideas arising in the program areas
• Need to develop an infrastructure for beta testing and
implementing new and better ways of doing things
• Collaboration with IT Directorate for computing needs,
and other directorates to ensure innovations are
understood and carried through
Strategic objectives 3
• Establishing more robust collaborations with
external researchers and agencies.
– Enhance collaborative research between R & M
personnel and outside researchers
– NSF/Census Research Network
– Expanded role for Research Data Centers
– Incentivizing internal and external researchers to tackle
methodological and subject matter research of high
value to the bureau and the broader economic and social
measurement communities.
– On topics of importance to the Bureau, defined
collaboratively with other directorates
Strategic objectives 4
• Increasing the statistical literacy of Census
Bureau data users
• Increased demand for products for small areas
(e.g. ACS)
• Increased “research” data releases that have
higher levels of error, are more model-based
• Users need to understand how to interpret
these data
Synopsis of NCRN
• A set of research nodes conducting interdisciplinary
research and educational activities on methodological
questions of interest and significance to both:
• The broad research community, and
• The Federal Statistical System, particularly the U.S.
Census Bureau.
• The initial awards established 2 small and 6 mediumsized research nodes; a later solicitation may permit the
establishment of large nodes.
• Complementary to but independent of the Census
Bureau’s Research Data Centers program.
Goals of NCRN
• Advance development of innovative methods and
models for the collection, analysis, and dissemination of
data in the social, behavioral, and economic sciences.
• Relate fundamental advances in methods development
to the problems of the Federal Statistical System,
particularly the U.S. Cen10sus Bureau.
• Facilitate the collaborative activities of scientists from
across multiple disciplines, including the social,
behavioral, and economic sciences, the statistical
sciences, and the computer sciences.
• Develop next generation of researchers in research skills
of relevance to the measurement of economic units,
households, and persons.
8 proposals funded in FY 2011
"Data Integration, Online Data Collection, and Privacy Protection for Census
2020". PI: Fienberg, Carnegie-Mellon.
"Integrated Research Support, Training and Data Documentation". PI: Abowd,
"Enhancing Federal Agencies' Data Dissemination Capabilities". PI: Reiter, Duke.
"Linking Surveys to the World – Administrative Data, the Web, and Beyond". PI:
Shapiro, Michigan.
"Improving the Interpretability and Usability of the American Community
Survey through Hierarchical Multiscale Spatio–Temporal Statistical Models". PI:
Holan, Missouri.
"Reducing Error in Computerized Survey Data Collection". PI: McCutcheon,
"Census Bureau Data Programs as Statistical Decision Problems". PI: Spencer,
"Improving the Connection between the Spatial and the Survey Sciences". PI:
Spielman, Colorado.
R and M Research Topics
• Census 2020 design
• Multiple mode data collection: Reducing
survey costs without sacrificing quality
• Disclosure avoidance
• Record Linkage, ad records
• Modeling to provide more frequent, more
granular estimates
• Innovative data products (combining
current data)
Some Challenges
• Recruit the best researchers
• Census Bureau employees, and outside researchers on
grants and contracts and as consultants
• Citizenship issues, competitive salaries, creating an
exciting research environment
• Value peer-reviewed publication, visibility in the wider
research community
• Promoting and fostering research excellence in a large,
production-oriented, bureaucratic institution
• Build better links between research and production
• Institutionalizing research excellence
We need your help!
• Let people know that the Census Bureau has
a new research directorate with exciting
• Give us your ideas
• ADRM wants to work with colleagues to develop
new and improved Census data products.
• We value your advice and constructive

similar documents