Proceedings of the Annual Meeting of the Cognitive Science Society

Age-related declines in scores on neuropsychological tests arewidely believed to reveal that human cognitive capacitiesdecline across the lifespan. In a computational simulation, weshow how the behavioral patterns observed in PairedAssociate Learning (PAL), a particularly sensitive measure ofage-related performance change (Rabbitt & Lowe, 2000), arepredicted by the models used to formalize associative learningprocesses in other areas of behavioral and neuroscientificresearch. The simulation further predicts that manipulatinglanguage exposure will reproduce the experience-relatedperformance differences erroneously attributed to age-relateddecline in age-matched adults. Consistent with this, olderbilinguals outperformed native speakers in a German PALtest, an advantage that increased with age. These analyses andresults show that age-related PAL performance changesreflect the predictable effects of learning on the associabilityof test items, and indicate that failing to control for theseeffects is distorting our understanding of cognitive and braindevelopment in adulthood.