The Development of Global and Local Processing: A Comparison of Children to Adults

In light of the adult model of a hemispheric asymmetry of global and local processing, we compared children (M _age = 8.4 years) to adults in a global-local reaction time (RT) paradigm. Hierarchical designs (large shapes made of small shapes) were presented randomly to each visual field, and participants were instructed to identify either the global or the local level in each of two blocks. We obtained evidence of a global-local processing asymmetry, with stronger effects for children than for adults. In both children and adults, responses were faster and more accurate for global identification in the right hemisphere and for local identification in the left hemisphere. Similarly, a significant asymmetry of global and local interference was obtained in children but not in adults. Interference reflects the RT cost of conflicting information at the nonattended level. For example, local interference indicates the degree to which inconsistency at the local level slows global identification. Stronger evidence for lateralized processing in children is discussed within the framework of increasing interhemispheric transfer. That is, as interhemispheric transfer increases, cost of presentation to the nonpreferred hemisphere (e.g., local identification in the right hemisphere) is reduced.     

Identity matching and oddity learning in patients with moderate to severe Alzheimer-type dementia

Discrimination learning was investigated in patients with moderate to severe Alzheimer-type dementia (AD), comparing their performance with age-matched controls. Four AD patients were trained to criterion on identity matching and then shifted to the same task with novel stimuli. The AD patients showed no savings in learning to match novel stimuli, whereas a comparative group of four control subjects rapidly learned the novel matching discrimination, maintaining criterion performance on the transfer test. A second group of four patients was initially trained on oddity, taking a similar number of trials to reach criterion as the matching group. When these patients were subsequently shifted to the matching task with novel stimuli, they performed substantially worse than the first group of patients who had learned the matching task in the first stage. The lack of positive transfer in the shift between matching to matching suggests that the AD patients solved the identity-matching task on the basis of stimulus-response associations rather than a rule. The presence of negative transfer after shifting from oddity to matching may be explained by a pre-disposition to respond to a novel stimulus that is carried over into the matching task, but this warrants further investigation, as indicated in the discussion.