Charting the trajectory of cognitive diagnostic assessment in language assessment: A systematic literature review

The education system has relied on conventional single-score assessments, which provide limited insights into students' cognitive abilities. Cognitive diagnostic assessment (CDA) offers a promising alternative by providing detailed information on students' cognitive processes. The review aims to explore the recent trends of cognitive diagnostic assessment publications in language assessment by examining the years of publications, geographical locations, research methodologies, focused skills, and research samples. It further investigates the implementation of CDA studies in language assessment and identifies proposed future research directions in this field. Findings show a surge in CDA studies since 2019, with China and Iran leading, predominantly focusing on receptive skills and retrofitting methods over true CDA. The review shows that CDA accentuates cognitive model design, Q-matrix construction, item development, comparison of cognitive diagnostic models (CDMs), data analysis, and mastery profile generation. Previous research has proposed recommendations related to item and instrument development, attribute selection, sample selection, and the effectiveness of CDA. The study limitations include potential biases from publication trends and the reliance on existing literature, possibly excluding unpublished works. Future efforts should focus on developing comprehensive cognitive models, balancing the assessment of receptive and productive skills, and expanding the analysis of item parameters. The review's findings imply that practitioners must align existing tests with diagnostic needs and incorporate various item parameters to enhance assessment accuracy. It also underscores the implication that educational policymakers should support refined cognitive models and that teachers must be trained to use advanced diagnostic tools for effective instructional strategies.