Confidence Intervals for a Mean and a Proportion in the Bounded Case

摘要

This paper describes a 100x(1-alpha) confidence interval for the mean of a bounded random variable which is shorter than the interval that Chebyshev's inequality induces for small alpha and which avoids the error of approximation that assuming normality induces. The paper also presents an analogous development for deriving a 100x(1-alpha) confidence interval for a proportion.