Medical Students’ Adoption and Perceptions of ChatGPT for EFL Academic Writing in Wartime Ukraine

• In Event: AERA Roundtable Session Sunday 11:45 am - Gold 3
  In Roundtable Session: Teachers’ Perspectives, Readiness and Experiences in the Post-COVID AI Era

Abstract

Using the extended Technology Acceptance Model (TAM3) and a mixed-methods design, this study examined how instructor-led training shapes EFL students’ perceptions, use, and evaluation of ChatGPT for academic writing amid wartime disruption at a Ukrainian medical university. Pre-intervention survey findings (N=191) revealed that, despite positive perceptions of AI's utility, students lacked formal guidance and expressed concerns about output accuracy and academic integrity. Post-intervention survey results (N=33) indicated high levels of satisfaction with the training, with qualitative responses revealing a shift from uncertain experimentation to more strategic, critical use of ChatGPT. Findings offer practical implications for curriculum design and teacher training in crisis-affected contexts and contribute a scalable model for integrating generative AI into language education where traditional instruction is disrupted.

Authors

• Haiyan Li, Purdue University

• Lixia Cheng, Purdue University

• Violeta Kalnitskaya, Purdue University