Data-driven learning in ESP university settings in Romania: multiple corpus consultation approaches for academic writing support

Abstract

Corpora are valuable technology-supported learning resources to be used by autonomous language learners or during teacher-guided lessons. This study explores the potential of corpus consultation approaches for the improvement of English for Specific Purposes (ESP) students’ academic writing skills. We investigated the effects of three types of Data-Driven Learning (DDL) activities in a sample group of 29 first-year and second-year students majoring in Geography for Tourism at a Romanian university, consisting of writing tasks supported by: a Learner Corpus (LC), a Native-Speaker Corpus (NSC), and a Web-based Corpus (WBC). The research methodology involves the combination of quantitative and qualitative data, extracted from pre- and post-intervention corpus analyses, with the results of a learner-satisfaction questionnaire. The findings indicate a significant differentiation in the complexity of the lexico-grammatical features used by learners in consequent intervention stages and a better integration of L2-related academic writing strategies into their written productions. The study yields first conclusions on the integration of computer-processed language databases in DDL strategies for ESP learners in the Romanian university context.