Rate of acquisition, adult age, and basic cognitive abilities predict forgetting: new views on a classic problem

Rate of forgetting is putatively invariant across individuals, sharing few associations with individual-differences variables known to influence encoding and retrieval. This classic topic in learning and memory was revisited using a novel statistical application, multilevel modeling, to examine whether (a) slopes of forgetting varied across individuals and (b) observed individual differences in forgetting shared systematic relations with adult age, learning speed, and cognitive ability. Participants (N = 136) received mnemonic training prior to memorizing 4-digit numbers to perfection, and retention was tested immediately after training and after 30 min, 24 hr, 7 weeks, and 8 months. Slower rate of learning to criterion, older age, and poorer cognitive performance predicted accelerated forgetting with associations most pronounced within 24 hr from baseline. Observed correlates of differential forgetting slopes are similar to those previously found to affect encoding, suggesting continuity rather than asymmetry of prediction for these memory processes.