Mediated Interaction and Advanced EFL Academic Writing: A Mixed-Methods Evaluation of a Sociocultural Online Writing Community

Abstract

Learning management systems may enhance second-language writing when their interactive affordances are integrated into purposeful pedagogical activities. This exploratory, qualitatively led mixed-methods case study examined how a Sociocultural Online Writing Community (SOWC), implemented through Blackboard, was experienced by advanced undergraduate English as a Foreign Language (EFL) students, and described the writing scores recorded in two intact sections of the same course (experimental, n = 30; comparison, n = 30). Over 15 weeks, the experimental section took part in structured online discussion, collaborative planning, rubric-guided peer feedback, and instructor mediation, whereas the comparison section used Blackboard mainly to access content, submit assignments, and receive private instructor feedback. Interviews with 15 experimental students, analyzed thematically, yielded six themes: dialogic idea generation, collaborative planning, peer-mediated revision, metacognitive awareness from evaluating peers, supportive participation, and design challenges. Students’ accounts indicate that asynchrony, the persistence of written feedback, and the public visibility of peers’ reasoning changed how assistance was offered and taken up, and that unequal participation and variable feedback quality limited it. The quantitative strand is descriptive. The experimental section began with higher pretest scores (19.17 vs. 17.30 out of 50) and recorded much higher posttest scores (44.73 vs. 17.33), and its students rated mediated interaction favorably (M = 21.43/25, α = 0.88). However, condition was confounded with section; the experimental section received about three times as many instructor comments per assignment and wrote on topics aligned with the posttest prompt; the essays were rated by members of the research team, including the course instructor; and the score distributions were atypically uniform in ways the authors could not explain from the retained records. The score difference therefore cannot be attributed to the SOWC design and is not interpreted as an effect estimate. The scores are reported in summary form only, pending editorial verification of the original posttest rater records and independent blind re-scoring of all 120 archived essay scripts by external raters, and statistics from earlier versions that could not be reproduced from the retained records have been withdrawn. The study’s contribution is a qualitative account of how peer, tool, and instructor mediation operated in an online writing environment, which can inform the design of confirmatory evaluations.