Recall performance of studied material typically increases across successive recall cycles, even in the absence of any corrective feedback. Prior research on this hypermnesia effect has predominantly employed free recall as the test format and has generally reported effects of moderate magnitude. Building on a recent incidental observation from a different experimental task, the present study examined whether an initial-letter cued recall (ILCR) format can amplify hypermnesia effects relative to a free-recall format. Across four experiments, participants repeatedly recalled previously studied item lists across three consecutive recall cycles. In Experiments 1-3, the test format involved either free recall (Experiments 2b and 3) or ILCR with a prescribed output order and a 6-s time limit per item (Experiments 1, 2a, and 3). Results showed larger hypermnesia effects in ILCR than in free recall, irrespective of whether words or images with their verbal labels served as study material and whether a 0-min delay or prolonged delays (30 min, 24 h) between study and test were applied. With both test formats, hypermnesia increased with delay. Experiment 4 showed that the results of Experiments 1-3 are critically tied to the constrained ILCR format and do not generalize to an ILCR condition in which all cues are presented simultaneously and participants are free to provide their responses in any order. Results align well with an interpretation based on both the changes in cue set account of hypermnesia and the retrieval practice hypothesis, suggesting a link between hypermnesia and the testing effect.
