Corpus-Based Materials Implementation in Writing Classrooms: Teachers’ Experiences, Perceptions and Training Initiatives

Abstract

Corpus linguistics has transformed language education by enabling teachers and learners to engage directly with authentic language data. Data-driven learning (DDL), also known as the direct use of corpora in language teaching (DDL; Johns, 1991), supports second language writing (SLW) instruction by fostering genre awareness and rhetorical decision-making (Boulton & Cobb, 2017). While research affirms the benefits of corpus-based instruction, most studies have centered on expert-driven models, limiting insights into how instructors without advanced corpus training learn to adopt these tools (Chambers, 2019). Moreover, many teachers receive little sustained support in integrating corpus methods into their instructional contexts (Farr, 2008; Poole, 2022). This dissertation investigates how teachers learn to design and implement corpus-based pedagogy in SLW instruction. Organized as a three-article study, it addresses gaps in teacher training, corpus-based materials design and implementation, and long-term engagement with DDL. Article 1 presents a systematic review of 33 empirical studies, identifying trends and gaps in DDL teacher education. Article 2 analyzes classroom data from 14 instructors implementing ready-made corpus-based materials, using the TPACK framework (Mishra & Koehler, 2006) to explore pedagogical decision-making. Article 3 examines a 10-week collaborative workshop in which 6 in-service SLW instructors engaged in co-design, peer feedback, and reflective dialogue around corpus pedagogy, highlighting how collaboration support teacher learning of corpus-based instruction. Together, these studies offer a multi-dimensional view of DDL integration and propose practical, collaborative, and context-responsive approaches to help teachers meaningfully incorporate corpus tools into SLW pedagogy. The opening chapter of the dissertation introduces the broader context of corpus linguistics, data-driven learning (DDL), and corpus-based instruction, while also presenting a summary of each article and my positionality as a researcher. This chapter is followed by three standalone articles (Chapters 2, 3, and 4). The dissertation concludes with a final chapter that synthesizes the insights across the three articles, highlights their collective contributions, and outlines implications and directions for future research.