The Effectiveness of a YouTube-Based Pronunciation Learning Model Integrating Local Wisdom for Junior High School Students

Abstract

It was aimed to assess the efficacy of integrating local knowledge into YouTube-based pronunciation instruction through a deep learning pedagogical framework. A Research and Development (R&D) design was utilized alongside a quasi-experimental method to create a culturally contextualized digital pronunciation module and assess its effect on students' pronunciation performance. The study took place at UPTD SMP Negeri 4 Bireuen in Aceh, with 60 students of the ninth grade randomly divided into an experimental group and a control group. We gathered data by giving them pronunciation tests before and after the tests and then used independent samples t-tests to look at the results. The preliminary test findings showed that there was no statistically significant difference between the two groups, which meant that their baseline pronunciation skills were similar. However, The subsequent test results indicated that the group performing the experiment significantly outperformed the control group (p < .05), which means that the instructional intervention worked. The results show that using Acehnese local wisdom, especially Khanduri Blang, in YouTube-based pronunciation learning helps students pronounce words more accurately, learn more consistently, and stay interested. The study theoretically enhances the application of deep learning as a pedagogical framework in culturally contextualized digital learning environments and provides practical implications for EFL educators and curriculum developers in creating meaningful and culturally responsive pronunciation instruction.