DOI: 10.12688/f1000research.188623.1 ISSN: 2046-1402

ENHANCING ASSESSMENT LITERACY: A SYSTEMATIC REVIEW OF AI TOOLS IN EDUCATION

Jesi Jecsen Pongkendek, Diana Rochintaniawati, Nahadi Nahadi, Eddy Prasetyo Nugroho, Wayan Suana, Andreas Suparman, Asep Saefullah, Wahyuni Satria Dewi, Mellyta Uliyandari, Alvyn Karina Lestari
Show PDF Cite
Abstract* Background Assessment is a core component of the learning process, and Artificial Intelligence (AI) is reshaping how educational assessments are designed, implemented, and evaluated. This systematic review synthesized evidence on the use of AI in educational assessment and examined its implications for strengthening assessment literacy. Methods A systematic search was conducted through Google Scholar to identify Scopus-indexed peer-reviewed articles published between 2017 and 2025. Studies focusing on AI applications in educational assessment were included, while non-peer-reviewed publications, opinion papers, and studies unrelated to assessment were excluded. Of 105 records initially identified, 46 studies met the eligibility criteria. Data were extracted and coded according to author characteristics, country of origin, research field, methodology, study population, AI-based assessment type, and research outcomes. Findings were synthesized using descriptive and thematic analyses. Results The included studies represented diverse educational contexts, involving participants ranging from children to graduate students and employing qualitative, quantitative, and mixed-methods designs. AI applications were identified across formative, summative, and diagnostic assessment practices. The findings suggest that AI can enhance assessment efficiency, support personalized learning, provide timely feedback, and facilitate data-driven decision-making. However, concerns regarding assessment validity, ethical issues, data privacy, and algorithmic bias remain persistent challenges. Conclusions AI shows considerable potential to improve educational assessment and strengthen assessment literacy among educators and learners. Nevertheless, the evidence base is limited by methodological heterogeneity, diverse educational contexts, and inconsistent reporting practices. Future research should develop robust, ethical, and evidence-based frameworks to guide the effective integration of AI into educational assessment systems.

More from our Archive