JEROME P. REITER
Assistant Professor of Statistical
Science
Department of Statistical Science
Duke University
Summary of
research interests
My first major
research project
involved methods of borrowing strength across groups without explicitly
pooling data across the groups.
This work
was motivated by an issue in the 2000
U.S.
census.
It
formed the basis of my Ph.D. dissertation, which is
summarized at
this linked page.
Since then, my
primary research focus
has been investigating statistical methods of preserving
confidentiality of data.
I'm currently
engaged in research on several topics, including (i) developing the
theory and assessing the feasibility of releasing synthetic, i.e.
simulated, data to the public; (ii) investigating the data security
and utility issues involved with
establishing
remote access servers that allow users to submit analyses and receive
output
without ever seeing the data; (iii) developing methods for
distributing data and analyses safely
across multiple statistical agencies; and (iv) developing methods for
assessing the disclosure risk and data utility of proposed disclosure
limitation strategies.
(Link to papers
on confidentiality research.)
More generally, I am
interested in research involving survey methodology, especially dealing
with missing data in complex surveys.
I
also am interested in research in causal inference and observational
studies. (
Link to papers on multiple
imputation, survey methodology, and causal inference.)
I very much enjoy using statistical
methods to analyze data and working with researchers who are not
statisticians.
I've analyzed data from business,
education, medicine, political
science, psychology, public health, and
sports. I'm always interested in new opportunities for
collaborations
in any discipline.
(Link to papers
for which the primary focus is data analysis.)
In one of my articles, I analyzed whether walking or pitching to Barry
Bonds was more effective. Check out some of the
newspaper stories
about my analysis.