Emily Grenner & Joost van de Weijer & Lena Asker-Árnason & Victoria Johansson & Viktoria Åkerlund & Birgitta S.M. Sahlén The aim of this intervention study is to investigate if observational learning can improve narrative writing skills in 11-year-olds with and without hearing impairment. Observational learning occurs when people learn new skills from observing others, who act as models (Bandura 1997). Observing peers’ reading and writing is especially important since these processes often are invisible, and children therefore lack models for their own processes. This study was theoretically and methodologically inspired by Rijlaarsdam et al. 2008.  Participants consisted of Swedish 5th-graders from two schools (School A, n=33; and School B, n= 26) with normal hearing children (NH), and from 3rd to 8:th-grade children with hearing-impairment (HI), from “hearing classes” (n=18). Prior to the intervention, background data e.g., on working memory and linguistic background was collected. In the research design the two schools with NH children (School A and B) functioned as each other's controls. The HI-school followed the School A order. All participants first wrote a personal narrative on the computer, using keystroke-logging. Then the intervention followed for School A and HI-school, while School B received ordinary lessons (with no writing instructions). After the first intervention period, all participants wrote a new narrative. Thereafter, the intervention was replicated for School B, while School A and the HI-school had ordinary tutoring. After the second intervention period, all participants wrote new narratives. The intervention consisted of 5 thematically different lessons: Lesson themes were: reader perspective, chronological structure, closing elements, revising of a peer’s text and online revision.  To evaluate the text quality, all texts (n=231) were holistically rated by three independent, trained evaluators. The results showed an improvement in quality between text 1 and text 2 for School A and the HI-School, while School B had an improvement between text 2 and text 3. This shows that narrative text quality can be improved by a short series of carefully designed intervention lessons using observational learning, which contributes to the discussion about educational methods for teaching writing.  Further analyses will address quantitative measures of text length, lexicon, syntactic complexity, pausing and editing, as well as a comparison between the NH and HI group.