Sequential study effects of free time and word frequency in long-term memory: A critical test of four theories using a large-scale publicly available dataset

A recent theory proposes that storing information in long-term memory depletes a limited resource which recovers gradually over time. Here, we provide a critical test of this theory by showing that free recall is worse if a word is preceded during study by a low-frequency word and by a short inter-stimulus-interval (ISI). These effects interacted – the preceding study word’s frequency effect decreased as the pre-ISI increased. The duration of subsequent ISI had no effect. The resource-depletion-and-recovery model explains these results by positing that the amount of resources required for encoding in long-term memory is an inverse function of the existing representation’s strength, and that these resources recover more when the pre-ISI is longer. The theory’s predictions are contrasted with three alternatives – temporal distinctiveness, selective rehearsal and refreshing, and short-term consolidation – none of which fit the data. We compare each set of predictions using the PEERS dataset (Healey & Kahana, 2016). The resource-depletion-and-recovery theory’s predictions provided the best match to the data.