Story composition and process analysis using a computer tool based system

The performance of children in writing is a cause for concern: national testing at age 11 indicates that too many pupils underachieve in writing. The purpose of the project outlined in this thesis was to design a computer tool which would assist children aged between 7 and 11 years achieve a higher standard in narrative writing. The project involved four phases of research. First, a review of the research literature and government documents relating to literacy was undertaken, to establish differences between good and poor narrative writing, differences in the approach of mature and beginner writers, the natural course of children's writing development, effective methods of teaching, ways of assessing writing quality and approaches adopted by other computer applications. Secondly, a computer tool which scaffolds the narrative writing process was designed. HARRY delivers conversational prompts, stored in Microsoft Access databases, in a conference like-sitaation. The tool acts as a bridge between the writing approach adopted by beginner writers and the mature approach whilst demonstrating features of successful narrative writing. Thirdly, a method for automating the assessment of children's written grammar was devised. CHECK TEXT, a computer utility program, provides quantitative analyses of specific grammar features, facilitating comparisons between examples of stories written at Key Stage 2. Both applications use ASP and HTML files, for implementation on a web server. Finally, two small scale studies were conducted in a primary school - the first to investigate the effects of HARRY's assistance upon children's writing performance and behaviour, and the second to investigate the effects of using the system upon children's subsequent writing performance. The first study showed that when supported by HARRY, the children adopted a more reflective writing approach and they produced more successful narratives, provided they followed the advice. The second study showed that HARRY can accelerate children's writing development - most of the children wrote subsequently better stories and they all felt that they had improved. All the children liked using the tool and thought that HARRY made writing stories easier. HARRY's limitations are indicated, and potential developments and further investigations are proposed.