<p>The integration of Big Data analytics into engineering education introduces transformative opportunities for personalized learning, performance assessment, and curriculum adaptation. However, this raises complex ethical challenges regarding data privacy, algorithmic fairness, transparency, and student autonomy. This survey presents a systematic and focused review of recent developments at the intersection of educational data mining and ethical accountability, with an emphasis on engineering contexts. Drawing on peer-reviewed studies published since 2020, the analysis maps the dominant analytical approaches, including predictive modeling, multimodal learning analytics, and reinforcement-based systems, against emerging concerns such as surveillance risks, bias propagation, and consent ambiguity. This study offers a comparative synthesis of real-world use cases, evaluates mitigation strategies, and identifies recurring gaps between practices and governance. The study concludes by outlining the pedagogical and societal implications and proposing directions for ethically informed system design and policy alignment. The present survey aims to support researchers, practitioners, and decision-makers in navigating the ethical landscape of data-intensive educational technologies.</p>