Assessment Model and Empirical Study of Students’ STEM Innovation Ability in Virtual Simulation Environment

To address the core issue of the “emphasis on operational accuracy but neglect of innovative thinking” and the lack of a scientific quantification model in the application of virtual simulation technology in STEM education, this study aims to construct an assessment system for STEM innovation ability that is compatible with the virtual simulation environment and to verify its effectiveness. The study first combed through the components of STEM innovation ability and the theories of virtual simulation teaching through literature research methods. Combined with the Delphi method (12 STEM education experts, technical engineers, and front-line teachers participated), it determined 12 assessment indicators across four dimensions: “problem discovery, solution design, interdisciplinary integration, and outcome optimization”. Subsequently, the Analytic Hierarchy Process (AHP) was used to calculate the weights of the indicators, and a quantification assessment model was constructed by integrating the Fuzzy Comprehensive Evaluation method. An automatic scoring system was also developed to achieve real-time conversion from “operational data to innovation scores”. Finally, 200 students from two middle schools (one urban and one rural) and one vocational college in Ningbo were selected to conduct a comparative experiment over one semester (virtual simulation group vs. traditional teaching group), and SPSS was used for reliability and validity tests and difference analysis.