Aging and False Memories: Comparing Effects of Item-Relatedness and List Position

Evidence from classic serial position curves suggests that semantic and phonological false memories traditionally arise from presumed long- and short-term memory stores, respectively. However, research challenges this distinction, finding semantic and phonological false memories across memory stores and list positions. The present study tested whether semantic and phonological false memories reflect distinct or shared mechanisms and whether these patterns differ with age.

Younger (YAs) and older (OAs) adults completed a modified Deese-Roediger-McDermott paradigm in both recall (Experiments 1 and 2) and item recognition (Experiment 3), manipulating semantic and phonological similarity across list position.

During recall, although OAs recalled fewer items than YAs, both groups generated more phonological than semantic false memories. During recognition, both groups had higher hit rates for phonological than semantic items and lower false alarm rates for the phonological compared to the semantic condition. Across all experiments, both groups made semantic and phonological false memories from all sublist positions, challenging notions of strictly separated memory stores. Importantly, irrespective of test format, there were no age-related differences in false memory patterns.

Together, these findings challenge traditional views of memory organization and suggest that shared cognitive processes across age group and list composition underlie semantic and phonological false memories.