Graduation Year


Document Type




Degree Granting Department

School of Aging Studies

Major Professor

Jerri D. Edwards


brain fitness, cognitive interventions, cognitive stimulation, cognitive training, memory, older adults


Background: With age, older adults experience declines in both short- and long-term memory. One way to counter these age-related declines is through memory interventions which include computerized cognitive training and non-computerized cognitive stimulation. This dissertation examined whether a cognitive training program, Dakim BrainFitness (Dakim Inc., 2002) and a program of cognitive stimulation, Mind Your Mind (Seagull & Seagull, 2007), enhance memory performance among cognitively-intact older adults residing in independent-living retirement communities. Specifically, the following research questions were proposed: (a) How effective is the computerized cognitive training program in improving memory performance relative to the cognitive stimulation program or a no-contact control condition? (b) How effective is the non-computerized cognitive stimulation program, Mind Your Mind, at improving memory performance relative to a control condition? and (c) Will memory training gains endure 3-months post-training for those who participate in cognitive training?

Method: Fifty-three older adults were randomized to cognitive training (n = 19), cognitive stimulation (n = 17), or a no-contact control (n = 17) condition. Participants in the cognitive training and cognitive stimulation conditions were asked to complete five 25-minute sessions per week for a 10-week period. Memory outcome measures included the Auditory Verbal Learning Test (AVLT), the Hopkins Verbal Learning Test (HVLT), and the Wechsler Memory Scale-Third Edition (WMS-III) Family Pictures subtest. Outcome measures were administered at baseline, immediately post-training (or equivalent delay), and again at 3-months post-training.

Results: Multivariate Analysis of Variance indicated no significant differences between the three training conditions on baseline characteristics and memory outcome scores (p = .660). To test hypotheses one and two, memory outcome measures were compared across training conditions and testing occasions. A repeated measures MANOVA indicated a significant group x time interaction, Wilks' Λ =.585, F(10,92) = 2.83, p = .004, partial η2 = .235. Follow-up analyses for each memory outcome measure from baseline to immediately post-training were conducted with training condition as the independent variable. Significant group x time interactions were found between conditions for AVLT delayed recall, F(2,50) = 3.683, p = .032, partial η2 = .128, and the HVLT immediate recall, F(2,50) = 5.059, p = .010, partial η2 = .168. No significant group x time interaction was indicated on the AVLT immediate recall, F(2,50) = 2.544, p = .089, partial η2 = .092. There was a marginally significant group x time interaction on the WMS-III Family Pictures delayed recall F(2,50) = 2.975, p = .060, partial η2 = .106.

Post-hoc comparisons for significant outcome measures were conducted using Fisher's LSD test, while controlling for baseline performance. Results indicated that the cognitive training condition performed significantly better than the cognitive stimulation condition from baseline to immediately post-training on the AVLT delayed recall (p = .012), as well as on HVLT immediate recall (p < .001). The cognitive training condition also performed significantly better from baseline to immediately post-training as compared to the no-contact control condition (p = .011). A significant difference between the cognitive training condition and the no-contact control condition was also found on the WMS-III delayed recall measure (p = .030) immediately post-training. No significant differences between any of the conditions were found on either AVLT immediate or WMS-III Family Pictures immediate recall (ps > .05). There were no differences between the cognitive stimulation and control conditions across all memory outcomes (ps > .05). For hypothesis three, a repeated measures MANOVA indicated no main effect of time within the cognitive training condition for the memory outcome measures, Wilks' Λ = .047, F(6,11) = 2.11, p = .135, partial η2 = .535.

Discussion: These findings provide evidence that the adaptive computerized cognitive training program, Dakim BrainFitness, significantly improved memory abilities as measured by the AVLT delayed recall, HVLT, and WMS-III Family Pictures delayed recall relative to cognitive stimulation. In contrast, there were no significant improvements for participants in the non-adaptive, non-computerized program of cognitive stimulation relative to controls. These findings coincide with the Model of Adult Cognitive Plasticity that in order to improve cognitive performance, there needs to be a mismatch between the individual's capacities and the demands of the task. Adaptive cognitive training may be more likely to provide a mismatch and produce positive plasticity changes in the brain. Future research pertains to exploring the cognitive benefits that these programs have on other types of cognitive domains.