Exploring Hellotalk in enhancing english-speaking proficiency among Indonesian EFL students: A mixed-methods study

Purnama, Tarmidzi Tibyan (2026) Exploring Hellotalk in enhancing english-speaking proficiency among Indonesian EFL students: A mixed-methods study. Masters thesis, UIN Sunan Gunung Djati Bandung.

Preview	Text (COVER) 1_cover.pdf Download (65kB) | Preview
Preview	Text (ABSTRACT) 2_abstract.pdf Download (142kB) | Preview
Preview	Text (SUKET BEBAS PLAGIASI) Surat bebas plagiarism Byan.pdf Download (178kB) | Preview
Preview	Text (DAFTAR ISI) 3_daftarisi.pdf Download (353kB) | Preview
Preview	Text (BAB I) 4_bab1.pdf Download (173kB) | Preview
	Text (BAB II) 5_bab2.pdf Restricted to Registered users only Download (526kB) | Request a copy
	Text (BAB III) 6_bab3.pdf Restricted to Registered users only Download (557kB) | Request a copy
	Text (BAB IV) 7_bab4.pdf Restricted to Registered users only Download (839kB) | Request a copy
	Text (BAB V) 8_bab5.pdf Restricted to Registered users only Download (154kB) | Request a copy
	Text (DAFTAR PUSTAKA) 9_daftarpustaka.pdf Restricted to Registered users only Download (189kB) | Request a copy
Preview	Text (APPENDIX) 10_appendix.pdf Download (6MB) | Preview

Abstract

Integrating Artificial Intelligence (AI) into English as a Foreign Language (EFL) instruction, offers new opportunities to tackle the ongoing challenges of speaking proficiency among Indonesian learners. This research reports the effectiveness of HelloTalk, an AI-supported mobile learning platform, in improving the speaking performance of undergraduate students at UIN Sunan Gunung Djati Bandung. Drawing on Vygotsky’s Sociocultural Theory, Krashen’s Input Hypothesis, and principles of Communicative Language Teaching, a mixed-methods approach was used. This included IELTS-style pre- and post-tests, post-intervention questionnaires, semi-structured interviews, and classroom observations. Quantitative data showed a significant improvement in speaking skills (p = .011) with a large effect size, along with consistently high perception scores (α = .943) across different demographic groups. Qualitative analysis offered deeper insights: learners expressed less speaking anxiety, greater willingness to communicate, and felt safer practicing with AI-mediated feedback. They valued the motivation from interactive exchanges, the convenience of mobile access, and the authenticity of practicing with global partners. Observations confirmed a shift from teacher-led discussions to more spontaneous, student-driven conversations and peer collaboration. However, challenges such as dependence on premium features, occasional tech issues, and exposure to inappropriate cultural content arose. Both students and teachers emphasized the importance of integrating these tools more structurally into the syllabus. This study demonstrates that with proper scaffolding and cultural responsiveness, AI tools like HelloTalk can bridge the gap between classroom instruction and real-world communication, offering scalable, interactive, and context-sensitive solutions for EFL education in Indonesia. These findings carry practical implications for educators, policymakers, and developers seeking to advance oral English proficiency in resource-constrained and culturally diverse learning environments.

Actions (login required)

View Item