Generative artificial intelligence as a pedagogical strategy to strengthen productive English skills in B1 level university students

Abstract

This research examines the pedagogical impact of integrating generative artificial intelligence (AI) tools—specifically ChatGPT, Grammarly, and QuillBot—on strengthening productive English speaking and writing skills in A2-level university students in Ecuador. The study employed a quantitative quasi-experimental design with two intact groups, experimental and control, totaling 150 participants. For eight weeks, the experimental group worked on structured oral and written tasks supported by AI applications, while the control group used traditional teaching methods. Data collection included pre-tests and post-tests assessed with rubrics based on the CEFR, as well as a Likert-type perception questionnaire administered to the experimental group. The results showed statistically significant improvements in the experimental group, with notable advances in coherence, fluency, grammatical accuracy, and vocabulary richness. Students with greater participation and frequent use of the tools demonstrated superior performance, confirming the added value of active engagement in technology-mediated learning. Student perceptions were mostly positive, highlighting the immediate feedback, improved organization of ideas, and increased autonomy; however, limitations were identified related to the need for constant teacher guidance and the risk of technological dependence. Taken together, the findings provide empirical evidence from a Latin American context, underscoring that, when applied within a sound ethical and pedagogical framework, generative AI tools have high potential to foster communicative competence, drive innovation in higher education, and prepare students for the linguistic and digital demands of the 21st century.

Ähnliche Arbeiten