| Abstrakt: |
The book "Confidence Intervals for Discrete Data in Clinical Research" by Vivek Pradhan, Ashis K. Gangopadhyay, Sandeep M. Menon, Cynthia Basu, and Tathagata Banerjee argues that significance testing is overemphasized and that data analysts should focus more on analyses using confidence intervals. The book provides a comprehensive overview of various methods for calculating confidence intervals for discrete data, including those based on asymptotic versions of tests, exact versions using small-sample discrete distributions, and small-sample methods with conditional and unconditional approaches. The authors also discuss frequentist and Bayesian approaches and provide examples and instructions for obtaining results using software such as SAS and StatXact. However, the book has some limitations, such as the lack of detail in certain methods and the omission of the odds ratio. Overall, it is a valuable resource for researchers interested in analyzing contingency tables and comparing the performance of different methods. [Extracted from the article] |
