Author

• Herman Chernoff

Abstract

The usual calculation of the P-value for the classical problem of‎ ‎comparing probabilities is not always accurate‎. ‎This‎ ‎issue arose in the context of a legal dispute which depended on when‎ ‎some written material was written in a diary‎. ‎The problem raises‎ ‎some issues on the foundations of statistical inference.‎ 

Keywords

• Ancillary‎
• ‎Bayes‎
• ‎binomial‎
• ‎Fisher exact test‎
• ‎likelihood-ratio‎
• ‎p-values‎
• ‎Yates continuity correction