Different patterns of recollection for matched real-world and laboratory-based episodes in younger and older adults.

To bridge the gap between naturalistic and laboratory assessments of episodic memory, we designed time- and content-matched real-world and virtualized versions of the same tour event. In younger and older adults, we investigated objective and subjective aspects of recollection for event features using a verbal true/false test common to both event conditions. Using a data-driven multivariate analysis blind to the age groups and event conditions, we found that discriminating altered from true details accounted for the largest amount of variance in objective retrieval patterns. There was an advantage for real-world over laboratory encoding on this dimension for both age groups. Similarly, real-world encoding elicited higher scores on a dimension defined by subjective recollection. However, real-world (but not laboratory) encoding decoupled objective and subjective memory in older adults, who reported similar rates of subjective recollection as younger adults despite exhibiting significantly poorer discrimination accuracy. These results demonstrate robust and specific ways in which the accuracy and subjective quality of memory differ for matched naturalistic and laboratory episodes. Furthermore, these results suggest that naturalistic and laboratory encoding conditions produce qualitatively different patterns of episodic memory decline in older age.