Combining dependent P-values

By J. Kost, M. Mcdermott. 2002

169 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

A General Method for Detecting Interference Between Units in Randomized Experiments (2012, unknown): “Note that while we do not make the requisite assumptions here, the p values for multiple subsets may be combined to obtain a single p value for the entire sample (e.g., Kost and McDermott 2002) if a covariance matrix for the p values is known.”
Understanding Membership Inferences on Well-Generalized Learning Models (2018, methodology): “enhancing records, we repeat the attack in section 4.4 by querying each of these records. Because the outputs on these queries may be correlated, we combine the resulting p-values using Kost’s method [23], with the covariance matrix estimated from the query features generated in the query selection ”
Combining p-values via averaging (2012, unknown): “at the p-values are independent, whereas we would like to avoid any assumptions besides all pk, k = 1,...,K, being bona ﬁde p-values. Fisher’s method has been extended to dependent p-values in, e.g., [1, 8], but the combined p-values obtained in those papers are approximate; in this note we are inte”