Abstract
Cognitive training programs for older adults often result in improvements at the group level. However, there are typically large age and individual differences in the size of training benefits. These differences may be related to the degree to which participants implement the processes targeted by the training program. To test this possibility, we tested older adults in a memory-training procedure either under specific strategy instructions designed to encourage semantic, integrative encoding, or in a condition that encouraged time and attention to encoding but allowed participants to choose their own strategy. Both conditions improved the performance of old-old adults relative to an earlier study (D. Bissig & C. Lustig, 2007) and reduced self-reports of everyday memory errors. Performance in the strategy-instruction group was related to preexisting ability; performance in the strategy?choice group was not. The strategy-choice group performed better on a laboratory transfer test of recognition memory, and training performance was correlated with reduced everyday memory errors. Training programs that target participants' latent but inefficiently used abilities while allowing flexibility in bringing those abilities to bear may best promote effective training and transfer.