Date of Award


Document Type



Edith Cowan University

Degree Name

Bachelor of Education Honours


School of Education


Western Australian College of Advanced Education

First Supervisor

Dr P. Sloan


In order to examine the effect of the teaching of a top-level expository structure through writing on children's schemata for text structure, children in year seven were asked to display graphically the relationships between 42 randomly presented text items related to a central topic. The text items were designed to approximate the kinds of information that may be found in an encyclopaedia or a science text about a given topic. One class of children was then allocated to the control group and the other to the experimental group. The experimental group was taught the top-level structure for a scientific report using a specific writing strategy. The control group were taught the top-level structure for the narrative using a similar basic strategy. After approximately four, seventy five minute, treatment sessions a post test was administered to determine if there were any changes in the complexity of the associations between given text items that the students were able to make. Three weeks later a third test was administered to determine if there had been any long term change to the students' text structure schema. A comparison of the performance of the experimental and control group in the post-test and delayed test supported a hypothesis that the treatment would cause long-term changes to the structure of an individual subject's semantic memory. The results also showed the limitations of teaching reading using only narrative materials. This study supported the research findings of Sloan (1983) which concluded that fluent readers differed significantly from less fluent readers in their ability to generate diagrams showing complex semantic relationships. The effectiveness of the treatment was also compared against previously established measures of reading fluency in order to determine if there were any correlations. An analysis of the data showed that the treatment was effective (p < .01) for two of the three categories of reading fluency established.