At first glance, the results of the latter trial appear promising, but the conclusions are clouded by two limitations. First, the study was not placebo-controlled, which is a major drawback given the possibility of motivational effects on cognitive test performance.[24] Second, the description of the training protocol, the cognitive tasks, and their relationship was vague, which makes it difficult to evaluate whether the measures displaying training-related effects are measures of transfer of training or directly pick up task-specific skills taught in the training. This issue must receive much greater attention in clinically oriented work on cognitive training. Notably, improvements in task performance are composed of different components, some of which (e.g., effective strategies and perceptual expertise) are not operationally defining improvements in the cognitive ability that the tasks are assumed to measure (e.g., episodic memory, processing speed, and working memory). Thus, the observed performance increments on the trained tasks do not necessarily imply that cognitive capabilities in and out of the laboratory have been improved. Transfer of training must be used to address this concern. Future research must apply broad batteries of transfer tasks and provide task analyses that clarify the relation between training and transfer tasks. Currently, it remains unclear whether intervention-related changes in intellectual engagement can have lasting effects on cognition in old age and reduce the incidence of dementia.