Evidence on the Effectiveness of Comprehensive Error Correction in Second Language Writing

268

Multilingual secondary-school pupils in Dutch classrooms writing in L2 Dutch - the youngest substantial sample in the written-corrective-feedback literature, and still adolescents rather than children

The strongest positive test of whether error correction transfers out of the corrected text and into new writing. Four conditions: direct comprehensive correction, indirect comprehensive correction, and two active control groups - self-editing without feedback, and additional writing practice without feedback. Having "more writing practice" as a control arm is what makes this study unusually informative, because it tests the opportunity-cost objection directly. Outcomes measured on revision of the corrected text AND on new texts written at posttest and delayed posttest (1 and 4 weeks later), with grammatical and non-grammatical error types analysed separately. Recorded from the published abstract. Population caveat: L2 Dutch in secondary school - nothing here licenses a claim about a native-speaking six-year-old.

Both direct and indirect comprehensive corrective feedback improved accuracy over self-editing without feedback AND over extra writing practice without feedback, "and this was true not only during revision but also in new pieces of writing" at 1 and 4 weeks. The error-type decomposition is the part that matters for a conventions topic: "only direct CF resulted in grammatical accuracy gains in new writing" while "pupils' NONGRAMMATICAL accuracy benefited most from INDIRECT CF" - the non-grammatical category being the one closest to mechanics. Correction did not make pupils write more simply: structural complexity and lexical diversity in new writing were unaffected, which falsifies the avoidance objection. The authors read the whole pattern as support for comprehensive correction as a durable teaching tool rather than a proofreading service.

Low. Pupils were allocated to four instructional conditions and every arm wrote and revised, so the contrast is between what the teacher did with the text, not between children who write and children who do not.