Construction and Management Application of a Big Data-Based Quality Evaluation System for Experimental Teaching in Science Education

Against the backdrop of the digital transformation of education, and aiming to address issues such as strong subjectivity and lack of process data in traditional science experiment teaching evaluation, this study constructs a quality evaluation system for science experiment teaching integrating four dimensions—students, teachers, resources, and management—from the perspective of educational management, based on big data technology. The system adheres to three principles: integration of process and outcome orientation, balance of objectivity and practicality, and dynamic adaptability, while clarifying the definition, data source, and quantification method of indicators for each dimension. Data are collected through experiment teaching platforms, smart devices, etc., and after undergoing a four-step preprocessing procedure, a layered and progressive analysis strategy (descriptive statistics, correlation and causation analysis, prediction and early warning) is adopted to explore data value. The evaluation results are applied relying on a three-tier feedback-optimization closed loop. A pilot application in chemistry experiment teaching at a university shows that the system increases device utilization rate by 15% and student operation compliance rate by 20%, effectively solving the pain points of traditional evaluation. Meanwhile, the study identifies challenges such as data governance and teachers' technical barriers, and puts forward optimization suggestions, providing decision support for educational administrators.