An investigation of the cognitive basis for the selectivity of age-related memory impairment

Older adults have been found to have a selective impairment in certain types of episodic memory, although other types of memory are generally preserved. The goal of this research is to determine whether the selective age-related memory deficit is best explained by an impairment in perceptual processing, an impairment in the formation of associations between items and their contexts, or an impairment in controlled processing, which is presumed to be required for recollection. Three behavioral experiments were conducted which attempted to evaluate the relative merits of each of these three accounts of age-related memory impairment. To allow for a more meaningful comparison of the data from each experiment, the same participants completed all three behavioral experiments. In addition to the behavioral experiments, an event-related potential (ERP) experiment was conducted to provide further information regarding perceptual processing differences between older and younger adults. When relying solely on perceptual information, rather than semantic and perceptual information, older adults' memory performance was especially poor for perceptually impoverished stimuli (words), but less so for perceptually rich stimuli (pictures). Unlike young adults, older adults did not benefit from repeated presentations of pair information, suggesting that older adults do not form associative links between to-be-remembered stimuli. However, older adults did not show a recollection-specific impairment as the controlled processing hypothesis would have predicted. Older adults were equivalently impaired for both recollection and familiarity measures, suggesting that controlled processing is not specifically impaired in older adults. ERPs for older adults had much more individual variability than for young adults and the differences in ERP waveforms between age groups were observed more consistently in word conditions than in picture conditions, which is consistent with the behavioral results. Additionally, older adult ERPs to pictures were most similar to young adults, in accordance with the behavioral results. The behavioral data support the hypothesis that there is a deficit in perceptual processing which may help explain age-related memory impairments. The ERP data, though limited, lends some support to this explanation as it reveals perceptual and semantic processing differences between young and older adults. An associative deficit may be an additional source of memory impairment