Changes in Seventh-Grade Students’ Writing Performance Following a Process Writing Intervention: An Exploratory Rasch Analysis

Abstract

This exploratory study examined changes in seventh-grade students’ English writing performance following a short process-writing intervention. A quantitative one-group pretest–posttest design involved 16 students (11 boys and 5 girls) from one Indonesian junior high school. Across five instructional sessions conducted over two weeks, students engaged in planning, mind mapping, drafting, feedback, revising, and editing. Writing was evaluated across six analytic traits—ideas, organization, grammar, vocabulary, conventions, and style/voice—and analyzed with a Rasch Rating Scale Model and paired-samples procedures. Mean Rasch person measures increased from −3.96 logits (SD = 3.51) at pretest to 2.28 logits (SD = 1.76) at posttest. The mean within-student change was 6.23 logits, 95% CI [4.27, 8.20], t(15) = 6.78, p < .001, dz = 1.69; a Wilcoxon sensitivity analysis also indicated higher posttest measures (p < .001). Person and item reliabilities were .92 and .91, respectively. Nevertheless, one student displayed substantial underfit, and the first residual contrast had an eigenvalue of 4.6 despite 68.9% variance explained by the Rasch measures, suggesting possible residual multidimensionality. The results indicate a pronounced within-cohort improvement after the intervention, but causal claims are not warranted because the study lacked a comparison group, used different prompts across occasions, and involved a small pilot sample. These findings provide preliminary evidence that structured process-writing instruction supported by iterative planning, feedback, and revision activities can be a valuable approach for developing adolescent EFL writing skills, while emphasizing the need for further validation through larger controlled studies.