Abstract
This study investigated the relationships of processing capacity and knowledge to memory measures that varied in retrieval difficulty and reliance on verbal knowledge in an adult life-span sample (N = 341). It was hypothesized that processing ability (speed and working memory) would have the strongest relationship to tasks requiring active retrieval and that knowledge (vocabulary ability) would be related to verbal fluency and cued recall, as participants relied upon verbal knowledge to retrieve category items (fluency) or develop associations (cued recall). Measurement and structural equation models were developed for the entire sample and separately for younger (aged 20–54 years, n = 168) and older (aged 55–92 years, n = 173) subgroups. In accordance with the hypotheses, processing ability was found to be most highly related to free recall, with additional significant relationships to cued recall, verbal fluency, and recognition. Knowledge was found to be significantly related only to verbal fluency and to cued recall. Moreover, knowledge was more important for older than for younger adults in mediating variance in cued recall, suggesting that older adults may use age-related increases in knowledge to partially compensate for processing declines when environmental support is available in memory tasks.
Acknowledgement
Trey Hedden is supported by an NRSA fellowship from the National Institutes of Health. This research was supported by grant R01-AG06265 from the National Institute on Aging to Denise C. Park. The authors thank Ki Goosens.
Notes
Significantly different covariances were between fluency and working memory, Δχ2 (1, N = 341) = 7.29, p = .007, and between vocabulary and fluency, Δχ2 (1, N = 341) = 6.97, p = .008. In both cases, the covariance was greater for the younger than for the older adults.
The positive relationship of age to vocabulary was in part due to the presence of a path from speed to vocabulary, which has a strongly positive value. Removing the path from speed to vocabulary greatly reduces the fit of the model, Δχ2 (1, N = 341) = 74.02, p < .001, but does not significantly alter any of the path values from vocabulary to the memory measures. This path suggests a strong relationship between speed and vocabulary, which may be expected due to the substantial relationships reported between speed and intelligence measures (e.g., CitationSalthouse, 1992). However, including this path also uncovers a substantial relationship between age and vocabulary, and it suggests that ageing is associated with two counteracting processes that affect vocabulary ability—lowered speed of processing tends to reduce access to vocabulary, while increased experience tends to increase vocabulary knowledge.