| dc.description.abstract | The rapid advancement of artificial intelligence has introduced new possibilities in language education, particularly in the assessment of speaking skills. This study investigates English language teachers’ perceptions, experiences, and evaluations regarding the use of ChatGPT as a tool to assess students’ speaking performance. Using a qualitative case study approach, data were collected through semi-structured interviews with three English language teachers in Pare, Indonesia, who have integrated ChatGPT into their speaking classes.
The results showed various insights from the participants regarding the use of ChatGPT as a supportive tool that can provide immediate feedback, generate speaking materials through prompts instruction, and help evaluate students’ grammar, vocabulary, and fluency. By implementing ChatGPT as speaking assessment, the participants described that students became more confidents expressing their ideas after receiving the feedbacks, especially in constructing more complex sentences. However, there are also some limitations that have been identified which can decline the effectiveness of ChatGPT as speaking assessment. It is found that ChatGPT is ineffective in assessing pronunciation, intonation, and non-verbal communication. Technical constrains are also concluded, such as limited access in the free version and the need for a stable internet connection.
Overall, this study viewed ChatGPT as an effective tool to help teacher in assessing speaking performance. However, it cannot completely replace traditional methods which involve human assessment and interaction due to ChatGPT’s inability to assess body and facial expression. Educators are advised to integrate ChatGPT carefully, combining its strengths with teacher-led assessments to ensure a thorough evaluation of speaking proficiency. | en_US |
