4.2 Article

Melded Confidence Intervals Do Not Provide Guaranteed Coverage

Related references

Note: Only part of the references are listed.
Article Mathematical & Computational Biology

Noninferiority testing with censoring when the event rate is low

Shannon K. Gallagher et al.

Summary: This study compares five methods for testing noninferiority in a setting with low event rates and finds that the normal approximation tests perform well, while the Kaplan-Meier methods have greater power, especially when there is high loss to follow-up.

STATISTICS IN MEDICINE (2022)

Article Mathematical & Computational Biology

Confidence intervals for difference in proportions for matched pairs compatible with exact McNemar's or sign tests

Michael P. Fay et al.

Summary: This article presents a method based on sign test for handling paired data and calculating effect estimates and confidence intervals. It proposes a confidence interval compatible with exact sign test, and supports the conjecture of guaranteed coverage through extensive numerical calculations.

STATISTICS IN MEDICINE (2021)

Article Statistics & Probability

Practical valid inferences for the two-sample binomial problem

Michael P. Fay et al.

Summary: This passage discusses the study of differences between two binomial parameters and their magnitudes, highlighting previous reviews and current research aspects. The author evaluates the compatibility of p-values and confidence intervals, showing that no one method can meet all the desirable properties.

STATISTICS SURVEYS (2021)

Article Ethics

SHOULD KIDS PLAY (AMERICAN) FOOTBALL?

Patrick Findler

JOURNAL OF THE PHILOSOPHY OF SPORT (2015)