INTEGRATING AUTHENTIC AND MODEL-BASED ASSESSMENT IN ENGLISH LANGUAGE LEARNING: A COMPREHENSIVE EVALUATION FRAMEWORK FOR HIGHER EDUCATION

Abstrak

Current debates in higher education English language assessment center on the tension between pedagogically valued authentic assessment and the growing demand for transparency, defensibility, and program-level accountability. While authentic tasks are widely promoted for representing real-world language use, their capacity to support valid inferences and systematic evaluation remains contested. This study engages with theories of Evidence-Centered Design, argument-based validity, and learning-oriented assessment to address this tension through an integrative program evaluation lens. Drawing on an empirical examination of authentic assessment practices in a higher education English program, the study demonstrates how model-based assessment theory can inform and strengthen current practice without undermining pedagogical authenticity. The proposed framework articulates key theoretical features, including explicit construct and evidence modeling, validity argumentation, and alignment across task, course, and program levels. Findings show that authentic tasks generate rich language performance evidence, but only become evaluatively actionable when embedded within explicit inferential and evaluative structures. The study concludes by arguing that integrating authenticity with model-based assessment and program evaluation enables assessment systems that are simultaneously learning-oriented, inferentially defensible, and institutionally usable, offering a principled pathway for best practice in higher education English language assessment.