Memory for truths and falsehoods

Three experiments were carried out to determine whether recognition memory was more efficient for true statements (on the basis of the subjects’ stable body of knowledge) than for false statements. The first two experiments did not show any difference. In a third experiment, the distractors were structured to prevent subjects from responding on the basis of the familiarity of individual component words rather than on the basis of the complete sentences. This experiment showed a clear difference between sentence classes, with true sentences remembered more accurately and faster than false sentences. The data fit, in general, the hypothesis that true sentences are doubly represented in memory. The relation of these findings to the findings on the effects of encoding tasks is also noted.