Related Experiment Videos
Patterns of non-response to a mail survey
C A Macera1, K L Jackson, D R Davis
1Department of Epidemiology, School of Public Health, Columbia, SC 29208.
Journal of Clinical Epidemiology
|January 1, 1990
Summary
This study on mail survey non-response bias found responders had more family history of chronic diseases and healthier behaviors. Both groups were equally healthy initially, but these differences impact future health outcome analyses.
Area of Science:
- Epidemiology
- Health Survey Research
- Biostatistics
Background:
- Non-response bias is a significant concern in mail surveys, potentially affecting the generalizability of health outcome findings.
- Longitudinal studies require robust methods to address participant attrition and ensure data integrity.
Purpose of the Study:
- To investigate potential non-response bias in a mail survey assessing health outcomes.
- To compare characteristics of responders versus non-responders in a longitudinal health study.
Main Methods:
- Comparison of personal health history, laboratory measurements, family disease history, demographics, health behaviors, and fitness metrics between responders and non-responders.
- Analysis of mail survey data from participants in a longitudinal study on physical activity, fitness, and health.
Main Results:
- Responders and non-responders showed similar personal health history and lab results at baseline.
- Responders reported a greater family history of cardiovascular disease, hypertension, and stroke.
- Male responders were younger, exhibited healthier behaviors, and had better weight and treadmill performance.
Conclusions:
- Initial health status was comparable between responders and non-responders.
- Individuals with a family history of chronic diseases and positive health behaviors were more likely to participate in the survey.
- Non-response bias can influence the interpretation of epidemiologic study results, highlighting the need for bias examination methods.