Independent paths in the development of infant learning and forgetting.

We investigated the possibility that age differences in infants' long-term retention are artifacts of correlated differences in learning rates or learning opportunities (over-learning). Using path analytic procedures, these possibilities were examined in two experiments in which 15- and 18-month-olds (Experiment 1) and 12- and 15-month-olds (Experiment 2) learned five novel activities to a strict acquisition criterion. Three months later, infants' retention was tested using four test trials with no further study opportunities. Using a series of causal models to test the relationships between age, learning rate, learning opportunities, and forgetting rate, the results disconfirmed the artifact hypothesis. These analyses indicated that, at least for criterion-learning designs, developmental declines in forgetting rates between 12 and 18 months of age do exist independent of developmental differences in learning. Furthermore, age differences in forgetting rates are not confounded with age differences in "overlearning." These findings are discussed in terms of the growing body of evidence that attests to the continuity of memory development across childhood.