Assessment and Practice in Educational Sciences 3092-717X 4 Serial Number 14 2026 04 13 Algorithmic Bias Awareness and Ethical Reasoning: Moderating the Impact of Systemic Biases in Generative Artificial Intelligence on Critical Thinking and Information Literacy in Higher Education Algorithmic Bias Awareness and Ethical Reasoning: Moderating the Impact of Systemic Biases in Generative Artificial Intelligence on Critical Thinking and Information Literacy in Higher Education 1 12 10.61838/japes.205 EN Journal Article 2025 09 17 <p>This study aimed to examine whether algorithmic bias awareness and ethical reasoning moderate the effects of perceived systemic biases in generative artificial intelligence on critical thinking and information literacy among higher education students in Tehran. The study employed a quantitative, cross-sectional correlational design with a moderation framework. Participants were 351 undergraduate and postgraduate students from public and private universities in Tehran selected through multistage cluster sampling. Data were collected using validated instruments measuring perceived systemic AI bias, algorithmic bias awareness, ethical reasoning, critical thinking, and information literacy. After preliminary data screening, hierarchical regression and structural equation modeling were conducted to test direct and interaction effects while controlling for demographic variables and frequency of AI use. Model fit was evaluated using standard goodness-of-fit indices. Perceived systemic bias in generative AI significantly and negatively predicted both critical thinking and information literacy. Algorithmic bias awareness and ethical reasoning each showed significant positive main effects on both outcome variables. Interaction analyses revealed significant moderation effects, indicating that high levels of algorithmic bias awareness and ethical reasoning substantially weakened the negative impact of systemic AI bias on critical thinking and information literacy. The structural equation model demonstrated excellent fit and confirmed the robustness of the proposed conceptual framework. The findings indicate that while systemic biases in generative AI pose measurable risks to essential academic competencies, these risks can be effectively mitigated through the development of algorithmic bias awareness and ethical reasoning, underscoring the necessity of embedding these capacities within higher education curricula and AI governance frameworks.</p> https://journalapes.com/index.php/apes/article/download/205/170