<![CDATA[The ability to speak fluently and appropriately in German remains a persistent challenge for Indonesian learners at the A2 level, especially in contexts that demand spontaneous, pragmatic communication. Many existing evaluation tools focus narrowly on grammatical accuracy, often neglecting fluency, lexical diversity, and sociopragmatic competence. This study aims to develop a corpus-informed speaking evaluation instrument specifically designed for assessing A2-level German learners in Indonesian higher education. Employing a Research and Development (R&D) framework using the ADDIE model (Analysis, Design, Development, Implementation, Evaluation), the instrument was systematically constructed through needs analysis, corpus-based task design, expert validation, and iterative refinement. Data collection involved interviews, classroom observations, and document analysis, supported by expert judgment and inter-rater reliability testing. The final product includes task-based prompts grounded in authentic communicative scenarios, a Likert-based rubric covering fluency, accuracy, vocabulary range, and pragmatic appropriateness, and lexical input informed by spoken German corpora. Results demonstrate the instrument's validity and reliability, while qualitative feedback supports its pedagogical relevance. This study offers a novel contribution to corpus-informed language assessment and serves as a model for localized speaking evaluation tools aligned with CEFR standards.]]>
Copyrights © 2025
