Expository Discourse Production in School-Age Children across Two Scaffolded Tasks.

School-age children encounter expository discourse daily in the classroom, and skilled understanding and production of expository language is critical for academic success. The purpose of this study was to compare children's production of two types of expository discourse, generation and retell, while employing a scaffolded note-taking procedure to assist children in developing their samples. Twenty-six typically developing children, 9 to 12 years of age, participated in the study. For the expository generation task, children gave an explanation of a favorite activity, and for the retell task they viewed a video and provided an explanation of the information in the video. Overall, expository generation samples were longer and richer in content, but expository retell samples demonstrated greater lexical diversity. In addition, generation samples contained fewer grammatical errors than retell samples (trend), but measures of syntactic complexity, which were positively related across tasks, were not significantly different between sample types. Findings suggest that using a scaffolded procedure for supporting expository production resulted in (1) samples that were long enough for valid analysis of the children's language and (2) the production of utterances that, on average, were longer than reported in similar studies without scaffolding. The potential impact of comprehension on retell task performance is discussed.