Continuity corrected score confidence interval for the difference in proportions in paired data

Area of Science:

• Biostatistics
• Statistical Methods

Background:

• Confidence intervals for risk difference with paired binary data often use hybrid or score methods.
• Existing asymptotic intervals may not adequately control coverage probability.

Purpose of the Study:

• To develop a novel score interval with continuity correction for improved risk difference estimation.
• To identify optimal correction values balancing coverage and interval width.

Main Methods:

• Proposed a new score interval incorporating continuity correction.
• Investigated three distinct correction values through simulation studies.
• Derived non-iterative solutions for the corrected intervals.

Main Results:

• A small continuity correction value demonstrated superior performance for the score interval.
• The proposed method achieved a better balance between coverage probability and interval width.
• Non-iterative solutions were successfully derived.

Conclusions:

• The developed continuity correction score interval enhances the reliability of risk difference estimation.
• The proposed method offers a practical and efficient alternative to existing intervals for paired binary data analysis.