Developing a Genre-Based Authentic Assessment Model for Academic Writing in Higher Education: An Indonesian Case Study

Abstract

Assessment in higher education has been shifting from traditional testing toward authentic performance-based assessment that mirrors real-learning experience. In the language-center industry, specifically composition instruction, this transition highlights contextualized and genre-conscious evaluation. Nevertheless, most institutions still rely on commercial-oriented grading practices, which are non-standardized, non-egalitarian, and not educative. In order to fill this gap, the study attempts to construct and validate a Genre-Based Authentic Assessment Model (GBAAM) for academic writing at the tertiary level in Indonesia. Using a 4-D design framework (Define, Design, Develop, Disseminate), the study embraced authentic assessment principles in a genre-based pedagogy to develop such a model: broad and context-sensitive. Findings revealed favorable content validity (CVI 0.82- 0.94), a strong level of reliability (α = 0.91), and a positive impact on students’ writing skills with a significant difference (p < 0.05) in their written proficiency from the upper-intermediate level in terms of dimensions such as structure, coherence, and contextual relevance. Qualitative evidence supported the model's utility and impacts on instructor objectivity and student engagement. The results indicate that the GBAAM successfully fills a gap between theory and practice by incorporating assessment for communicative competence and cultural context. The findings in this study provide a transferable model for academic writing assessment and propose future developments in digital or AI-assisted rubrics.