UBC Theses and Dissertations
Age differences in the semantic structure of animal terms and the effects of training Storm, Jill Christine
Two studies were conducted to explore the role of semantic structure within the domain of animal terms in children's spontaneous operations on these terms and on learning and transfer performance. In the first study, 24 subjects at each of 6 educational levels from kindergarten through zoology graduates were given four tasks to perform on a set of animal terms. Associative tasks (free listing and animal associations to animal words) showed similar semantic structures across all educational levels, with a few strongly associated terms, and a large group of fairly isolated terms. While a size dimension was dominant in the sorting and similarity tasks for grade 7, grade 11, and undergraduate groups, the dominant dimension for zoologists was based on food habits. In the second study, 140 grade 3 and grade 7 subjects learned a list of familiar animal terms. List organization was experimentally varied. Two types of organization were both designed to train the same multiple classification scheme: a hierarchically organized presentation of the items together with their category labels; and a list presentation of the items designed to train separately the class inclusion relations involved in the hierarchical structure. A random organization of the stimulus items was also included. The experimental hypothesis was that for older children the hierarchical organization would be as effective as the separate presentation of the two classifications, while for younger children the separate presentation would be more effective. Free recall was the direct measure of learning. A series of four sorting tasks with the same item set were used to test for transfer. Based on the Piagetian notion of increasing cognitive flexibility in children throughout the concrete operational period, it was expected that older children from both organized training conditions would combine and disembed dimensional values on the transfer task more accurately and spontaneously than younger children. A no-training control group at each age level performed on the transfer task. Older children recalled more items. Organized presentation produced superior recall in early trials to randomly organized lists at both ages. Type of organization did not produce differential effects on recall, nor did this variable interact with age. Measures of clustering in free recall were also analyzed. There was a training effect at both grade levels on the free sorting task. Groups receiving structured organization had essentially learned a new and different way of organizing these lexical items. Grade 7 children applied the structures learned more spontaneously and with fewer errors than younger children in the same conditions. On the 6, 3, and 2 category sorts, grade 7 children were more accurate than grade 3 children. Trained subjects performed better than non-trained subjects on the 6 and 3 category sorts. On all sorts hierarchical and list condition subjects were more accurate than random condition subjects. Grade 7 children in the organized stimulus conditions did better on all sorts than their grade 3 counterparts. Hierarchical condition subjects did better than list condition subjects on the 6 category sort, but equivalently on the 3 and 2 category sorts. The theoretical and practical implications of these results are discussed.
Item Citations and Data