The Impact of data-driven learning on the improvement of grammatical proficiency in the ESL classroom environment

Abstract

Corpus-based data-driven learning (DDL) is an innovative approach that utilises electronic text collections for linguistic analysis, thereby enhancing teaching practices and learning skills for ESL/EFL students. This innovative method surpasses traditional language teaching approaches. The aim of this study was to examine the impact of incorporating corpora in teaching grammatical constructs, specifically subject-verb agreement rules for ESL/EFL students. Furthermore, it examined ESL/EFL students' perceptions of using corpora for grammar instruction. A mixed method research design was employed, collecting both quantitative and qualitative data through triangulation methods. Data collection involved written essays, two timed writing tasks (a pretest and posttest), and an individual semi-structured interviews. Quantitative data were analysed using rubrics and paired samples t-tests, while thematic analyses were applied to the qualitative data. The analysis of the essays revealed that the students made errors in subject-verb agreement. The paired sample t-test revealed a statistically significant p-value (.001<0.05), indicating a notable improvement in the students' mastery of  subject verb agreement rules after receiving DDL instruction. In addition, qualitative interview responses indicated that participants held positive opinions about learning through the DDL approach. They described it as enjoyable, fascinating, and challenging, and believed it to be an effective method for acquiring new grammatical skills. The study concluded with recommendations for English Language Teaching (ELT) instructors and curriculum designers.