Exact One-Sided Confidence Limit for the Ratio of Two Poisson Rates
Statistics in Biopharmaceutical Research
First page number:
Last page number:
This article examines exact one-sided confidence limits for the ratio of two independent Poisson rates. The Buehler method is used to obtain exact limits, and this method is used in conjunction with existing approximate limits. The method of variance estimates recovery (MOVER) is a general approach to construct the ratio of two independent Poisson rates from the confidence intervals of each rate which can be obtained from commonly used methods. Four existing approximate limits are considered: the Wilson interval, the MOVER Jeffreys interval, the MOVER Rao score interval, and MOVER Rao score interval on log-scale. The exact limits respect the coverage requirement, and they are as small as possible under certain mild conditions. Our numerical studies indicate that exact upper limits using the MOVER Jeffreys interval, and the MOVER Rao score interval have good performance. © 2017 American Statistical Association.
Exact One-Sided Confidence Limit for the Ratio of Two Poisson Rates.
Statistics in Biopharmaceutical Research, 9(2),