Impact of an extreme observation in a paired samples design

Abstract

The effect of systematically altering the value of a single observation within a paired differences design is considered. A paradox is observed for the paired samples t-test, where increasing the value of an observation in the direction of the true mean difference results in a higher p-value. Using simulation, deviations from robustness of the paired samples t-test is demonstrated, and is contrasted with Yuen's paired samples t-test and the Wilcoxon signed rank sum test.