Discovering formulaic language through data-driven learning: Student attitudes and efficacy

Corpus linguistics has established that language is highly patterned. The use of patterned language has been linked to processing advantages with respect to listening and reading, which has implications for perceptions of fluency. The last twenty years has seen an increase in the integration of corpus-based language learning, or data-driven learning (DDL), as a supporting feature in teaching English as a foreign / second language (EFL/ESL). Most research has investigated student attitudes towards DDL as a tool to facilitate writing. Other studies, though notably fewer, have taken a quantitative perspective of the efficacy of DDL as a tool to facilitate the inductive learning of grammar rules. The purpose of this study is three-fold: (1) to present an EFL curriculum designed around DDL with the aim of improving spoken fluency; (2) to gauge how effective students were in employing newly discovered phrases in an appropriate manner; and (3) to investigate student attitudes toward such an approach to language learning. Student attitudes were investigated via a questionnaire and then triangulated through interviews and student logs. The findings indicate that students believe DDL to be a useful and effective tool in the classroom. However, students do note some difficulties related to DDL, such as encountering unfamiliar vocabulary and cut-off concordance lines. Finally, questions are raised as to the students’ ability to embed learned phrases in a pragmatically appropriate way.