大模型支持下的教师教学胜任力伴随式测评研究  

Accompanying Assessment of Teachers’ teaching Competence Supported by Large Model

作　　者：胡小勇[1] 涂果果 谢雅淇 Xiaoyong HU;Guoguo TU;Yaqi XIE(Research Institute of Educational Artificial Intelligence,South China Normal University,Guangzhou 510631,Guangdong;School of Educational Information Technology,South China Normal University,Guangzhou 510631,Guangdong)

机构地区：[1]华南师范大学教育人工智能研究院,广东广州510631 [2]华南师范大学教育信息技术学院,广东广州510631

出　　处：《中国教育信息化》2025年第3期31-40,共10页Chinese Journal of ICT in Education

基　　金：2024年度国家社会科学基金一般项目“教师数字胜任力伴随式智能测评研究”(编号:BCA240050)。

摘　　要：推进教师教学胜任力高质量发展,关键要发挥好教师数字化教学评价的指挥棒作用。传统的教师教学胜任力评价方式存在主观性强、准确度低、测评维度不全面、数据利用率不高、缺乏个性化评估等不足。据此,研究面向教师教学胜任力的伴随式测评方法,为教师提供教学胜任力的实时追踪、反馈和优化,具有重要意义。首先梳理教师教学胜任力的内涵及其测评发展,分析当前教育领域中大模型技术的应用现状。然后基于研究背景,结合大模型技术和已有的实践研究,构建垂直大模型支持下的教师教学胜任力伴随式测评框架。该框架包含四个关键组成部分:测评维度选择、大模型基座设计、精准子模型构建和多模态数据表征,通过创新教学胜任力智能测评与画像构建方式,实现教师教学胜任力的个性化培养与发展。在此基础上,进一步提出测评大模型的四大应用场景:一是教学胜任力画像生成与更新;二是人机协同支持教师教学胜任力持续发展;三是优质资源与教师画像的精准匹配;四是面向教师发展的循证决策与管理。To promote the high-quality development of teachers’teaching competence,it is crucial to play the role of the commanding rod in teachers’digital teaching application evaluation.Traditional methods of evaluating teachers’teaching competence have many shortcomings,such as strong subjectivity,low accuracy,incomplete evaluation dimensions,low data utilization rate,and lack of personalized evaluation.Therefore,it is necessary to study the accompanying evaluation method for teachers’teaching competence,providing real-time tracking,feedback,and optimization of teachers’teaching competence.First,the connotation of teachers’teaching competence and its evaluation development were summarized,and the current status of big model technology application in the field of education was analyzed.Second,based on these research backgrounds,combined with big model technology and existing practical research,a vertical big model-supported teachers’teaching competence accompanying evaluation big model framework was constructed.The framework includes four key components:evaluation dimension selection,big model pedestal design,precise sub-model construction,and multi-modal data representation,to innovate the way of intelligent evaluation and profile construction of teachers’teaching competence,and realize personalized cultivation and development of teachers’teaching competence;Finally,four major application scenarios of the evaluation big model were proposed:the generation and update of teachers’competence profile,human-computer collaboration to support teachers’continuous development,precise matching of high-quality resources and teachers’profiles,and evidence-based decision-making and management for teachers’development,aiming to provide reference for relevant research.

关 键 词：大模型 多模态数据 伴随式测评 教师教学胜任力 教师画像 

分 类 号：G434[文化科学—教育学]