In this thesis, we take the most popular normal approximate interval for the proportion of success and improve on it so that its confidence coefficients are close to the pre-set confidence levels. In addition to the central limit theorem, we apply the Binomial-to-Poisson convergence theorem and the fiducial interval concept. In this thesis, we take the most popular normal approximate interval for the proportion of success and improve on it so that its confidence coefficients are close to the pre-set confidence levels. In addition to the central limit theorem, we apply the Binomial-to-Poisson convergence theorem and the fiducial interval concept.
