Testing the Identical Effect on Predicted and Actual Memory Through Pictorial Stimuli

People tend to predict better memory for identical word pairs (e.g., DOG-DOG) than related ones (e.g., DOG-CAT), despite remembering related pairs more accurately-a phenomenon known as the identical effect. Across three experiments, we examined whether this illusion extends to pictorial materials and investigated the roles of processing fluency and a priori beliefs. Participants studied image pairs that were identical, exemplars, related, unrelated, or rotated (in Experiment 3). After each pair, they made judgments of learning (JOLs), and memory was later tested by a cued four-alternative forced-choice (4-AFC) recognition test. Consistently, identical image pairs received higher JOLs than related ones, despite equivalent or poorer recall. Identical pairs were also identified more quickly, reflecting greater processing fluency. However, identification speed did not consistently predict JOLs, suggesting that processing fluency alone cannot explain the illusion. These findings indicate that both processing fluency and beliefs influence JOLs, with beliefs about the pair types playing a central role.