Author's School

Graduate School of Arts & Sciences

Author's Department/Program



English (en)

Date of Award

January 2009

Degree Type


Degree Name

Doctor of Philosophy (PhD)

Chair and Committee

Martha Storandt


Recent research in cognitive aging has brought renewed interest to a decades old question. Can cognitive skills be trained, and if so, how widely does that trained skill transfer? Previous research has demonstrated that older adults are able to improve their performance on laboratory cognitive tests and in some cases these benefits can transfer to other similar tests: e.g. Kramer et al., 2004). A few cases have demonstrated transfer to more distal outcomes: Willis et al., 2006). This area of research is still in an early stage, and reports are mixed with regard to the efficacy of cognitive training. These mixed findings are obscured by differences in methodology and construct conceptualization. For example, many studies do not include a control condition, fail to consider practice effects as an account for posttest gains in test performance, or use outcome measures that lack a validated association with the targeted intervention. This dissertation tested the transfer of cognitive training using an experimental approach that addressed previous limitations. A circumscribed set of cognitive variables representing processing speed, working memory, response distractor inhibition, and task switching were chosen to examine reliability, construct validity, and to characterize the training task. Training-related gains in performance were demonstrated in a group of healthy older adults. Transfer of training-related gains was greater than practice-related improvement observed in a control group on a novel, but similar task. Transfer to more remote cognitive variables associated with the training task were examined and indicated a possible relationship between training and cognitive processing speed. An important methodological aspect of this dissertation was the demonstration of the need to incorporate factor analytical approaches into the study of construct-level phenomena: e.g., attention), although large samples and high task reliability are prerequisite conditions for this approach and can be difficult to obtain. Further studies are needed to delineate the constructs associated with observed training gains and to specify the type and extent of transfer of training-related skills.


Permanent URL: