ELICITATION OF KNOWLEDGE TRANSFORMATIONAL REPORTS WHILE CHILDREN WRITE NARRATIVES

Bereiter and Scardamalia (1987) proposed a model of children's writing , labeled ''knowledge-telling,'' which they characterized as being sim ple and linear in nature. By contrast, their model of more expert writ ing is described as ''knowledge transformational,'' because it is more sophisticated in its involvement of complex problem-solving processes . These researchers report that children, and even most adolescents, d o not write at a level that would implicate knowledge-transformational processes, especially in expository prose writing. Their evidence con trasts with the reports of process-based writing researchers (such as Graves, 1985; 1991), who through ethnographic investigations have desc ribed youthful writing as more representative of transformative proces ses than the ''knowledge-telling'' model might predict. Alternative re search approaches are needed to evaluate these conflicting claims rega rding the developing writing process. A stimulated-recall procedure, w hich enabled protocol analyses to be performed on participants' in sit u comments, was employed with 53 twelve-year-old children to examine t heir reports of their own writing processes. A narrative-writing task was used as a context for these children in which to comment on their writing as they wrote. The protocols of the sessions were recorded, tr anscribed, and evaluated for evidence of transformative problem-solvin g behaviors (largely as identified by Bereiter & Scardamalia, 1987), i ncluding: extended start-up time, time-monitoring, task completion tim e, note-making, oral reports of planning, audience awareness, and revi sion of story plans, and the use of a range of sources for story ideas . Many of these sixth-grade participants showed evidence in their prot ocols of at least one, and some, as many as five of the targeted trans formative behaviors. Gender effects included performance quality, and time monitoring: Girls wrote higher quality stories, and boys monitore d time more during writing. Writing quality, in turn, related to plann ing and revision indices, which correlated with each other, and planni ng related to audience awareness. Revision of story ideas and start-up times were correlated. Of the transformative behaviors, planning and note-making significantly predicted writing quality. This pattern of f indings confirmed the presence in oral reports of certain transformati ve or recontextualizing (Cameron, Hunt, & Linton, 1996) processes duri ng the writing, at least during the narration, of these young writers.