This project identified, analyzed, tested, and presented other methodological options that could fill persistent data gaps for small racial and ethnic minority and other hard-to-reach populations in order to more effectively address health disparities impacting these groups. The project began with a systematic review, interpretation, and synthesis of the research literature on population-based surveys to identify and evaluate the extent to which methods of data collection/analysis have been successfully used to collect/analyze data on racial and ethnic minority populations. A cost-effective and statistically reliable approach to deriving estimates of prevalence of various health behaviors, risk factors, and outcomes in small racial and ethnic and other hard-to-reach groups was identified and developed, followed by validation of the proposed approach and identification of its limitations and the conditions under which it is best used.
The study resulted in the modification of an existing statistical tool (the Kalman Filter) and the development of the Modified Kalman Filter, a new statistical tool that uses national data to pool estimates over time to make valid inferences about small populations. The project included the development of software and a step-by-step manual to facilitate use of the Modified Kalman Filter by analysts and improved accuracy of health measures and outcomes for small racial and ethnic minority and other hard-to-reach population groups.
Report Title: Improving Modified Kalman Filter Procedures: Final Report
Agency Sponsor: OASH, Office of the Assistant Secretary for Health
Federal Contact: Audrey Burwell, 240-453-8222
Performer: Rand Corporation
Record ID: 8231 (Report issued April 29, 2011)