LOCAL IDENTITY, GLOBAL VOICE: IMPROVING SPEAKING SKILLS OF ESL LEARNERS THROUGH THE CULTURALLY RESPONSIVE-AUDIO DESCRIPTION (CRAD) MODEL

Abstract

This mixed-methods study investigated the efficacy of the Culturally Responsive Teaching- Audio Description (CRAD) model in improving the speaking skills of undergraduate ESL learners. Grounded in the Affective Filter Hypothesis (Krashen, 1985) and Culturally Responsive Teaching (CRT), the research hypothesized that leveraging local Pakistani heritage content would reduce speaking anxiety and provide a linguistic scaffold for complex oral production. The quantitative (experimental) phase (N=60) used a pre-test/post-test design, measuring fluency, linguistic complexity (T-unit analysis), and coherence. Results demonstrated statistically significant improvements for the CRAD group (p < .05) across all metrics, particularly in descriptive vocabulary and complex syntax. The subsequent qualitative phase analyzed student reflections (N=20), yielding themes of Reduced Speaking Inhibition (lowered anxiety), Targeted Linguistic Precision (necessitated use of complex structures), and Motivation and Refinement. The findings affirm that CRAD effectively bridges affective safety and linguistic complexity. The study concludes that leveraging students' local identity transforms the ESL speaking task into an empowering communicative act, offering a data- backed pedagogical model for advanced oral skill enhancement.