<![CDATA[Background: Language assessment is widely used to determine learners’ proficiency; however, the accuracy of score interpretation depends on the quality of the measurement instrument. Aims: This study aims to evaluate the psychometric quality of the Arabic Language Fluency Assessment (ALFA) using Classical Test Theory. The psychometric findings are subsequently interpreted from the perspective of communicative language assessment informed by the CEFR framework. Methods: A quantitative descriptive-evaluative design was employed using Classical Test Theory (CTT). Data were collected from 92 beginner-level learners and analyzed using item difficulty, discrimination index, reliability (Cronbach’s Alpha), and distractor functioning. Results: The findings show that the test is dominated by easy items, with mean difficulty indices of 0.93 and 0.88, resulting in limited score variability. Consistent with this pattern, most items exhibited poor discrimination, while a large proportion of distractors were non-functioning. Although the reliability coefficients were relatively high (0.885 and 0.843), these values indicate internal consistency rather than overall measurement quality. Implications: Within the present sample, the ALFA demonstrated limited ability to differentiate learner performance. The psychometric findings also raise the possibility of construct underrepresentation when interpreted from the perspective of communicative language assessment informed by the CEFR framework. The findings support systematic revision of the ALFA instrument and may also inform the improvement of institution-developed Arabic language assessments in comparable educational contexts.]]>
Copyrights © 2026
