Imagination inflation with posttest delays: how long will it last?

Abstract

When attempting to recall past events we often imagine what might have been. There is growing evidence that those acts of imagination may lead to memory for events that never happened. Imagination inflation occurs when the subjective probability that an event occurred increases after it has been imagined. Imagination inflation is typically demonstrated using a pretest-posttest design. A familiarity-attribution model for imagination inflation is proposed. The model predicts that as memory for what was imagined decreases the ability to attribute familiarity also decrease and this produces larger amounts of imagination inflation. In this study the time between the imagination activity and the posttest was 1, 7, or 14 days. Higher levels imagination inflation were found at the longer delays. When subjects recalled what they imagined pretest to posttest changes were not different across the 3 delays. When subjects did not recall what they imagined the difference between their pretest and posttest ratings increased at a steady rate from 1 to 7 and 14 days. The familiarity-attribution model accounts for these effects.