The University of Michigan Center for Excellence in Health Statistics seeks to improve the quality of survey data available to federal, state, and local policy makers for health promotion and disease prevention.  Survey methods have been called one of the great inventions of the 20^th Century, and are widely used in health policy formulation for information about the health and health care utilization of the general population.  Sound, accurate, and reliable data are required for policy analysts not only to assess the current state of health and health care utilization in the general population but also to develop models to predict trends and developments in the population’s health.  Improving the quality of health survey statistics leads to more effective health promotion and disease prevention policy.  The Michigan Center for Excellence in Health Statistics (MiCEHS) is designed to conduct research on the survey process to improve the accuracy of data obtained from sample surveys.

The Center builds on research at the University of Michigan and its partner in the proposed Center, the Gallup Organization, which uses quantitative and qualitative methods to study the survey research process.  Survey research methodology is at present a nascent discipline, investigating the survey process through interdisciplinary studies blending theory from social, behavioral, health, economic, statistical, and computer sciences into comprehensive theories of information collection and utilization.  The MiCEHS will develop theory to explain behavioral and statistical features of the survey process, and generate empirical findings to test and refine the theory.

  The MiCEHS has four principal aims:  

1.   Elaborate the existing infrastructure to support interdisciplinary survey methodology research in a setting where methodological practice and substantive application are joined.  

2.   Conduct interdisciplinary methodological research on the survey research process using, when appropriate, ongoing survey data collection activities at the University of Michigan and the Gallup Organization as vehicles for research and laboratory studies.

3.   Develop new statistical methods for analyzing survey data in collaboration with other researchers in the School of Public Health, the Gallup Organization, and the National Center for Health Statistics.    

4.   Promote educational outreach to health statistics researchers at the National Center for Health Statistics and elsewhere to inform about the latest research findings and to train in research methodologies in the survey research process.  

Infrastructure

The MiCEHS includes several initiatives to promote the research activities by Center members.  Two of these, an interdisciplinary seminar in health statistics and an NCHS data center, are discussed here.

An ongoing seminar series provides an intellectual meeting ground of Michigan researchers and visitors who are making contributions in fields related to health statistics.  The seminar meets every other week throughout the academic year (September through May) and is organized around themes on important issues facing the field of health statistics.  Themes can be methodological, soliciting participation of substantive researchers who are interested in methodological developments as contributors or discussants, or substantively motivated by measurement problems in health statistics.  A video link between Ann Arbor and other locations is used to broaden the participation to include National Center for Health Statistics staff.  

A second component of the MiCEHS infrastructure is an NCHS data center.  NCHS has developed an administrative mechanism to increase access to confidential data resources through a data center in its Hyattsville offices.  The MiCEHS is seeking NCHS approval to establish an offsite data center as part of the MiCEHS in order to increase the range of topics that the proposed MiCEHS can address.  For example, one of the three projects developed later will require access to data that are not part of the public use data released by NCHS at present.  

This infrastructure is designed to supplement the primary research activities of the Center.  The MiCEHS has begun with three research projects.  But a mechanism of renewal and innovation is needed to sustain a long-lasting contribution to health statistics.  A mechanism has been established within the Center to review, approve, and fund future research projects.  A system of review and approval jointly between the MiCEHS staff, NCHS staff, and a Center advisory committee will assure that the best research ideas are funded.  There will be preliminary development to assess the feasibility of alternative research projects among MiCEHS senior staff, proposal review, scoring by advisory committee and NCHS staff, and evaluation and selection of projects for funding.

Publication: 

Survey Methodology Research for Improved Health Statistics