Related Experiment Videos
Sample size requirements for interval estimation of the odds ratio
1Department of Biostatistics, School of Public Health, University of North Carolina, Chapel Hill 27599-7400.
American Journal of Epidemiology
|January 1, 1990
Summary
This study provides a method for calculating the smallest sample size needed for case-control or cohort studies to estimate the odds ratio within a specified interval and confidence level. The findings differ from previous methods based on expected cell counts.
Area of Science:
- Epidemiology
- Biostatistics
Background:
- Accurate sample size calculation is crucial for the validity of case-control and cohort studies.
- Interval estimation of the odds ratio is a key objective in these study designs.
Purpose of the Study:
- To develop a procedure for determining the minimum sample size for interval estimation of the odds ratio.
- To ensure a specified confidence interval width with a given probability.
Main Methods:
- The proposed method calculates the smallest sample size for a 100(1-alpha)% confidence interval of the log odds ratio not exceeding a specified width.
- Presents tables of sample sizes for diverse parameter values.
Main Results:
- The developed procedure yields the minimal sample size meeting specified precision and confidence requirements.
- Identified considerable disagreement with existing methods relying on expected cell counts.
Conclusions:
- The presented method offers a robust approach to sample size determination for odds ratio interval estimation.
- Highlights limitations of prior methods based on expected cell counts, suggesting a need for re-evaluation.