PENGEMBANGAN MEDIA PEMBELAJARAN INTERAKTIF AI-INTEGRATED MENGGUNAKAN AUDIO-LINGUAL METHOD UNTUK MENINGKATKAN PRONUNCIATION DAN SPEAKING SELF-EFFICACY MAHASISWA

DEVELOPMENT OF AI-INTEGRATED INTERACTIVE LEARNING MEDIA USING AUDIO-LINGUAL METHOD TO IMPROVE STUDENTS' PRONUNCIATION AND SPEAKING SELF-EFFICACY

        Penelitian ini bertujuan untuk mengembangkan dan mengevaluasi media pembelajaran interaktif berbasis AI dengan pendekatan Audio-Lingual Method (ALM) guna meningkatkan kemampuan pronunciation dan speaking self-efficacy mahasiswa. Penelitian menggunakan pendekatan research and development model ADDIE yang dilanjutkan dengan desain kuasi-eksperimen pada mata kuliah Speaking and Listening dengan melibatkan kelas eksperimen dan kelas kontrol. Media pembelajaran mengintegrasikan teknologi Speech Recognition dan Text-to-Speech (TTS) untuk mendukung latihan berulang dan umpan balik otomatis. Data dikumpulkan melalui tes pronunciation, kuesioner speaking self-efficacy, validasi ahli, dan respons pengguna, kemudian dianalisis menggunakan uji statistik parametrik dan non-parametrik. Hasil penelitian menunjukkan bahwa media pembelajaran yang dikembangkan sangat layak dan secara signifikan meningkatkan kemampuan pronunciation serta speaking self-efficacy mahasiswa dibandingkan pembelajaran konvensional, sehingga efektif mendukung akurasi linguistik dan kesiapan afektif dalam pembelajaran bahasa Inggris.

This study aimed to develop and evaluate an AI-integrated interactive learning medium based on the Audio-Lingual Method (ALM) to improve students’ pronunciation ability and speaking self-efficacy. Using a research and development approach with the ADDIE model followed by a quasi-experimental design, the study involved an experimental and a control group in a university Speaking and Listening course. The developed media integrated Speech Recognition and Text-to-Speech (TTS) to support repetitive practice and immediate feedback. Data were collected through pronunciation performance tests, speaking self-efficacy questionnaires, expert validation, and user responses, and analyzed using parametric and non-parametric tests. The results showed that the media was highly feasible and significantly improved students’ pronunciation performance and speaking self-efficacy compared to conventional instruction. The findings indicate that AI-integrated media grounded in ALM effectively supports linguistic accuracy and affective readiness in English language learning.