The effects of data-driven learning activities on EFL learners' writing development.

Area of Science:

• Corpus linguistics
• Second language acquisition
• Educational technology

Background:

• Data-driven learning (DDL) is effective for addressing writing challenges like errors and idea generation.
• Chinese learners of English frequently use search engines like Baidu for reference.
• BNCweb offers a user-friendly interface for corpus-based language learning.

Purpose of the Study:

• To empirically evaluate the effectiveness of data-driven learning with BNCweb for Chinese EFL learners.
• To compare the impact of BNCweb versus the search engine Baidu on writing skills.
• To explore learner attitudes and challenges associated with using corpora in writing.

Main Methods:

• An empirical study involving 48 Chinese college students learning English as a foreign language.
• Quantitative comparison of writing performance (fluency, accuracy, complexity) between a BNCweb group and a Baidu group.
• Qualitative analysis of learner interviews to understand attitudes and identify challenges.

Main Results:

• The BNCweb group demonstrated significantly higher writing fluency and accuracy in post-tests compared to the Baidu group.
• No significant difference was observed in writing complexity between the two groups.
• Learners generally expressed positive attitudes towards BNCweb, but noted challenges in corpus integration.

Conclusions:

• BNCweb enhances writing fluency and accuracy more effectively than Baidu for Chinese EFL learners.
• Corpus-based learning shows promise but may require supplementary resources to overcome barriers.
• Integrating corpora with other reference tools can optimize the learning experience for EFL students.