Abstract
Several hypotheses regarding the ability of elderly subjects to learn a fluid intelligence skill, problem-solving, were investigated. Elderly and young adult subjects were assigned to one of four training groups; a No Treatment Control Group, a Training-in-Test-Taking Control group, a Modeling group, or a group trained with Discovery Method techniques. Training took 4-8 hours (self-paced) and included a variety of problem-solving tasks. Dependent measures included two different types of problem-solving tasks to investigate generalization of the training; and two forms of each task, one relevant to daily life and one irrelevant, to investigate the effects of relevance. Attention was paid to methodological issues raised in previous research including motivation, length of training, pacing, and subject and situational variables.The results failed to support any of the hypotheses drawn from the literature. There were no significant differences between age groups in learning or the effects of relevance. The interaction of type of training and type of task more than any other factor, appeared to be the major determinant of learning. Implications of these findings as well as several possible explanations for the outcome of the study were discussed. It was suggested that subjects should be matched on problem-solving ability and that learning is enhanced when the type of training used teaches the skills necessary for successful completion of the dependent measure, though specific training on the task is not required. Most importantly, this study suggests that fluid intelligence skills can be taught to elderly subjects as well as to young adult subjects and that learning does not differ significantly between age groups.