This explanatory mixed method quasi experimental study tested a novel pedagogical protocol—Contrastive Self Modeling with Multimodally Annotated Authentic Interviews (CSM MAI)—that integrated feedforward self modeling, multimodal annotation, and side by side visual comparison. Eighty seven advanced male EFL learners (mean age 23.4) from the Alvand Institute in Qazvin, Iran, were randomly assigned to CSM MAI (n = 29), an authentic model only group (AMG; n = 29), and a scripted materials control (CG; n = 29). Instruments measured speaking anxiety, interactional self efficacy, and interactional competence (IC; holistic rubric + discourse features). ANCOVA revealed that CSM MAI led to significantly greater reductions in anxiety (partial η² = .41), larger gains in self efficacy (partial η² = .44), and superior IC scores (partial η² = .38) relative to both comparison conditions. Qualitative data, collected through semi structured interviews and analyzed with inductive thematic analysis within a qualitative descriptive approach, confirmed that the side by side contrast normalized disfluencies and enabled precision goal setting. The study contributes a replicable protocol grounded in social cognitive theory and demonstrates how visually mediated self to model comparison can reshape advanced learners' standards for interactional success. From a critical applied linguistics perspective, the protocol challenges the native‑speaker ideal by reframing disfluency as a normal, strategic feature of authentic communication.
Type of Study:
Research |
Subject:
General Received: 2026/06/1 | Accepted: 2026/10/26 | Published: 2026/10/26