Asesmen HOTS Bahasa Arab Berbasis Bukti Penalaran dalam Lingkungan Digital

Abstract

Higher-order Arabic assessment should elicit defensible reasoning rather than infer cognitive complexity from long texts or difficult vocabulary. This article develops an evidence-oriented framework for digital assessment through a purposive narrative review and analysis of constructed tasks. Selected Arabic assessment studies are connected with revised Bloom categories, argument-based validation, and formative assessment. The synthesis distinguishes the intended claim, the task that elicits evidence, the learner response, and the interpretation of the resulting score. A short Arabic announcement illustrates how the same stimulus can support factual retrieval, inference, evaluation, or text revision depending on the question and scoring criteria. Digital delivery is positioned as a way to preserve explanations and revisions, not as a guarantee of higher-order thinking. The framework proposes an assessment blueprint, a two-stage response format, and an initial analytic rubric separating answer accuracy, textual evidence, and inferential justification. It also identifies vocabulary burden, ambiguous alternatives, technological barriers, and feedback timing as potential threats to score interpretation. No learner data, item statistics, or validity coefficients are reported. The contribution is an untested design framework that requires expert review, response-process investigation, independent scoring, and classroom piloting before it can support decisions about learners’ higher-order Arabic abilities.