生成式人工智能在教育评价中的定位与实践OA

As educational demands in the digital-intelligence era shift toward large-scale personalized learning,standardized test-based evaluation is gradually giving way to diversified,generative evaluation.Generative artificial intelligence(GenAI)grounded in dialogue aligns closely with this transformation and demonstrates strong potential to support the entire process of educational evalua-tion.Based on recent practices in related fields,this study conceptualizes the positioning of GenAI in educational evaluation into three progressive stages from a metaphorical perspective of"dialogue":as an assisting technology for"dialogue"with evaluation materials,as an augmentative technology for"dialogue"within the evaluation process,and as a mediating technology for"dialogue"with evalua-tion subjects.On this basis,this study constructs an overarching framework of GenAI's positioning and practice,and explains the modes,forms,and effects of"dialogue"across different stages.The study not only offers theoretical insights into the the underlying mechanisms through which GenAI empowers educational evaluation,but also provides practical guidance for integrating GenAI into e-ducational evaluation contexts in a stratified and classified manner.