从经验到数据：AIGC驱动下职业本科辅导员思政评价模型的构建逻辑与指标设计

From Experience to Data: The Construction Logic and Indicator Design of an AIGC-Driven Ideological and Political Education Evaluation Model for Vocational Undergraduate Counselors

摘要: 职业本科辅导员思政教育评价长期面临指标同质化、过程性数据采集困难、评价结果滞后三重困境。为破解此难题，本研究构建了一个以人机协同理论为基石、以CIPP模型为框架的评价模型。模型采用“感知–分析–决策”三层架构，利用生成式人工智能(AIGC)技术处理学生非结构化行为数据，并创新性地引入多层线性模型(HLM)的层级分析思想，关注“班级–个体”的双重变异。通过德尔菲法与层次分析法(AHP)，构建并确定了涵盖政治素养、职业品格、工匠精神、集体意识、技术伦理5个一级指标、13个二级指标、22个三级指标的权重体系。以某职业本科院校2025级机电一体化专业三个班级(N = 126)为样本进行实证检验，结果表明：该模型能显著提升评价的区分度(评分标准差扩大，p < 0.05)，实现对学生潜在风险的早期预警，并量化揭示班级层面管理环境对学生思政素养的显著影响(ICC = 15.3%)。研究证实，AIGC可作为辅导员的“智能辅助”，在严格的数据伦理边界与“机器筛证、人工断案”的人机协同模式下，有效赋能思政教育评价的精准化与科学化。

Abstract: The evaluation of ideological and political education conducted by counselors in vocational undergraduate institutions has long been plagued by three intertwined challenges: homogenized evaluation indicators, difficulty in collecting process-oriented data, and lagging evaluation results. To address these issues, this study constructs an evaluation model grounded in human-machine collaboration theory and framed by the CIPP model. The model adopts a three-tier “perception-analysis-decision” architecture, employs generative artificial intelligence (AIGC) to process students’ unstructured behavioral data, and innovatively incorporates the hierarchical analysis logic of the hierarchical linear model (HLM) to capture “class-individual” dual variation. Through the Delphi method and analytic hierarchy process (AHP), a weighted indicator system is established, comprising 5 first-level indicators (political identity, professional character, craftsmanship spirit, collective consciousness, and technological ethics), 13 second-level indicators, and 22 third-level indicators. An empirical test is conducted with three classes (N = 126) of the 2025 cohort majoring in Mechatronics at a vocational undergraduate institution. Results show that the model significantly improves evaluation discrimination (standard deviation of scores expands, p < 0.05), enables early warning of potential student risks, and quantitatively reveals the significant impact of class-level management environment on students’ ideological and political literacy (ICC = 15.3%). The study confirms that AIGC can serve as an “intelligent assistant” for counselors, effectively empowering the precision and scientific nature of ideological and political education evaluation under strict data ethics boundaries and a human-machine collaborative mode of “machine screening evidence, human making judgments.”