Related Experiment Videos
Assessment of coverage rates and bias using double sampling methodology
Paul Jenkins1, Charles Scheim, Jen-Ting Wang
1Bassett Research Institute, One Atwell Road, Cooperstown, NY 13326, USA. paul.jenkins@bassett.org
Journal of Clinical Epidemiology
|May 6, 2004
Summary
Double sampling significantly improved health outcome prevalence estimation in a rural New York survey. This method reduced bias and increased generalizability, providing more accurate health data for public health initiatives.
Area of Science:
- Public Health
- Epidemiology
- Survey Methodology
Background:
- Health status surveys are crucial for understanding population health.
- Accurate prevalence estimation is vital for effective public health interventions.
- Nonresponse bias can significantly distort health outcome data.
Purpose of the Study:
- To assess the impact of double sampling on health outcome prevalence estimation.
- To evaluate the generalizability and bias reduction of double sampling methodology.
- To compare health outcome prevalence between initial responders and incentivized nonresponders.
Main Methods:
- A double sampling strategy was employed in six rural central New York counties.
- Prevalence estimates were calculated for initial responders, incentivized nonresponders, and the combined group.
- Neyman's double sampling methodology was utilized.
Main Results:
- Demographic variables were similar between groups, but disease and health behavior prevalences differed significantly.
- Failure to use double sampling would have overestimated chronic disease prevalence by up to 6.2% (females) and 2.0% (males).
- Double sampling enhanced the generalizability of results from ~35% to over 70% of the population.
Conclusions:
- Double sampling substantially decreases bias in health endpoint estimation.
- This methodology significantly improves the representativeness of survey findings.
- Double sampling is an effective technique for enhancing the generalizability of health surveys in rural populations.