The Effects of Peer-Assisted Sentence-Combining Instruction on the Writing Performance of More and Less Skilled Young Writers.

Fourth-grade students, more skilled and less skilled writers, assigned to peer-assisted sentence-combining or a peer-assisted grammar-instruction comparison

Grade 4 students (roughly ages 9-10) in US elementary schools

Small randomised comparison (Journal of Educational Psychology 97(1), 43-54) in which the CONTROL condition is itself grammar instruction delivered in the same peer-assisted format - which is what makes it useful. Most sentence-combining studies compare against business as usual; this one holds delivery format, dose, peer structure and teacher attention constant and varies only whether the content is sentence combining or grammar. Outcomes are researcher-scored writing measures. Sample size is small by METHODOLOGY's standard (n < ~100), so grade C.

Peer-assisted sentence-combining instruction outperformed the peer-assisted grammar comparison on sentence-combining skill and on story writing measures, with the pattern differing between more skilled and less skilled writers. The design's value for this archive is the head-to-head - when sentence combining and grammar instruction are delivered by the same method, to the same children, for the same time, the sentence-level construction work is what moves writing. Effects on revising were more limited.

Low - randomised assignment within a single grade cohort. The skilled/less skilled moderation is a within-arm split on a substantially heritable trait and should be read as descriptive, not as a demonstrated aptitude-treatment interaction.