This week's speaker at the Gov 3009 workshop is Mary Beth Landrum from
Harvard Medical School. She will be speaking on "Methodologic Issues
in Profiling Medical Providers."
Abstract:
Quantitative performance monitoring of health care providers has become an
increasingly prominent activity in health care delivery. The resultant
"report cards" allow researchers, policy makers, and consumers to evaluate
their health care providers. In this talk, I will present two-case
studies illustrating the use of statistical methods to advance the utility
of profiles and to examine the impact of provider profiling on patient
care. First, while increased access to quality data has been urged by
many as a fundamental step in medical care reform, the amount of reported
data on multiple dimensions of care can often be confusing and
overwhelming to policy makers and consumers alike, suggesting the need for
methods to summarize care. In the first case study we use latent variable
models to create profiles of mental health care provided to US veterans
using multiple indicators of care. Several analytic issues will be
discussed, including the multidimensional nature of the data, the desire
for summary measures of performance, the need for models appropriate for
mixtures of continuous and discrete outcomes, the clustering of patients
within providers, and difficulties in selecting a framework for selecting
poor performing providers. Second, despite the building momentum for
increased public reporting of quality data, the underlying premise of
profiling - that disseminating information about provider quality will
lead to improved quality of care - has yet to be established. In this
case study, we use a longitudinal data set containing all individuals who
underwent CABG in a hospital in New York State over a 6-year period
(1991-1996) to consider how report cards affect the behavior of individual
providers. Specifically we examine if providers who are publicly
identified as being significantly better or worse than their peers
experience greater changes in volume, average severity and mortality
following surgery compared to those not so labeled.
Seminar Information:
The seminar meets at noon in Room 22, Center for Basic Research
in Social Sciences (CBRSS, 34 Kirkland St., this is the yellow building
across the street from William James Hall). Contact information, previous
presentations, and the spring schedule may be found at the course web
site: http://www.courses.fas.harvard.edu/~gov3009/. Lunch is provided.
To join the gov3009 mailing list, send e-mail to
gov3009-l-request(a)fas.harvard.edu with the following text message:
subscribe
end
Questions? Please contact the workshop coordinator, Liz Stuart, at
stuart(a)stat.harvard.edu
This week's speakers at the Gov 3009 workshop are David Lazer and Nancy
Katz from the John F. Kennedy School of Government. They will be speaking
on "Putting the network into teamwork."
More information can be found at the website www.ksg.harvard.edu/teamwork/
Seminar Information:
The seminar meets at noon in Room 22, Center for Basic Research
in Social Sciences (CBRSS, 34 Kirkland St., this is the yellow building
across the street from William James Hall). Contact information, previous
presentations, and the spring schedule may be found at the course web
site: http://www.courses.fas.harvard.edu/~gov3009/. Lunch is provided.
To join the gov3009 mailing list, send e-mail to
gov3009-l-request(a)fas.harvard.edu with the following text message:
subscribe
end
Questions? Please contact the workshop coordinator, Liz Stuart, at
stuart(a)stat.harvard.edu
This week's speaker at the Gov 3009 workshop is Scott Desposato from the
University of Arizona. He will be speaking on "Correcting for Bias in
Roll-Call Cohesion Scores." An abstract is below, and a link to the full
paper can be found at: http://www.u.arizona.edu/~swd/rb.pdf
Abstract:
Cohesion scores are a standard measure of legislative behavior and common
in legislative studies. Their strengths explain their wide use: cohesion
scores are simple, intuitive, and easy to calculate. Perhaps because of
this, little attention has been paid to cohesion scores' statistical
properties or relationship to existing theories of legislative behavior.
In this paper, I show how under a basic random utility model of
legislative behavior, cohesion scores suffer a serious bias problem:
scores are artificially inflated for small parties and weak parties. This
bias challenges a consistent finding in the comparative literature on
political parties; that small parties are consistently more disciplined
than large parties. I propose an intuitive solution that will eliminate
bias for a wide variety of models: making large parties smaller
by drawing samples without replacement from them.
Seminar Information:
The seminar meets at noon in Room 22, Center for Basic Research
in Social Sciences (CBRSS, 34 Kirkland St., this is the yellow building
across the street from William James Hall). Contact information, previous
presentations, and the spring schedule may be found at the course web
site: http://www.courses.fas.harvard.edu/~gov3009/. Lunch is provided.
To join the gov3009 mailing list, send e-mail to
gov3009-l-request(a)fas.harvard.edu with the following text message:
subscribe
end
Questions? Please contact the workshop coordinator, Liz Stuart, at
stuart(a)stat.harvard.edu
This week's speaker at the Gov 3009 seminar is Donald Rubin of Harvard's
Department of Statistics. He will be speaking on "Fabry's disease and
randomized experiments for FDA: The roles of historical controls and
multiple imputation for serious causal inference." No paper is available
at this time.
Seminar Information:
The seminar meets at noon in Room 22, Center for Basic Research
in Social Sciences (CBRSS, 34 Kirkland St., this is the yellow building
across the street from William James Hall). Contact information, previous
presentations, and the spring schedule may be found at the course web
site: http://www.courses.fas.harvard.edu/~gov3009/. Lunch is provided.
To join the gov3009 mailing list, send e-mail to
gov3009-l-request(a)fas.harvard.edu with the following text message:
subscribe
end
Questions? Please contact the workshop coordinator, Liz Stuart, at
stuart(a)stat.harvard.edu