On the generality of the revelation effect.

Seven experiments demonstrate the robustness of the revelation effect, which is the tendency to call recognition test items old if they are distorted when they initially appear and if they are revealed before the recognition judgment. With anagrams as the distortion, a revelation effect was found in within- and between-subjects designs, in a frequency-judgment task, in a list-discrimination task, when new items were used as targets, when the study list and the test were presented in different modalities, and when the word that was revealed did not match the word that was recognized. These results challenge accounts that attribute the revelation effect either to an increase in the familiarity of the revealed test word or to a positive response bias.