Marzano’s taxonomy-guided AI-generated answer detection and analysis of student achievement in interactive classrooms

Bibliographic Data

ID21631601
AuthorsShumin Liu (0000-0003-2880-3254, Yunnan Normal University), Xiyang Li (0009-0003-4895-0494, Yunnan Normal University), Xunxian Li (0009-0001-8119-0920, Yunnan Normal University), Jia Hao (0000-0003-3730-0528, Yunnan Normal University, corresponding author)
Year2026
Pages1-14
Publication date2026-02-26
Open AccessNo
TypeARTICLE
VenueInteractive Learning Environments (JOURNAL)
Journal identifiersISSN: 1049-4820 • E-ISSN: 1744-5191
PublisherInforma UK Limited (PUBLISHER • GB)
DOI10.1080/10494820.2026.2635540
OpenAlexW7133869216
LanguageEN
References cited26

With Generated AI wide-adopted in higher education, students increasingly collaborate with AI for classroom interactions. However, the impact of such AI-assisted behaviors on learning outcomes remains unclear, largely due to the difficulty in accurately distinguishing AI-generated content from human-written responses in real-time. To address this, we focuses on the interactive classroom, proposing an AI-Generated Answer Detection framework (AIGD) which integrates Marzano's Taxonomy of Educational Objectives with deep learning methods. Specifically, we first classify teachers’ classroom questions based on Marzano's Taxonomy, and then embed the results into the detection model to introduce semantic features from different cognitive levels. Second, a prompt-learning-based data augmentation method with the pre-trained ALBERT+CNN deep learning model is employed to balance the distribution of AI-generated and human-answered samples to mitigate the sample imbalance. Finally, guided by Marzano's Taxonomy, we analyze the correlation between students’ propensity to use AI at different cognitive levels and their academic performance. The empirical analysis reveals that AI usage acts differently across cognitive levels, which provides a new perspective for understanding how AI usage impacts learning outcomes, ultimately supporting teachers’ design of questioning strategies

Academic achievement · Computer-Assisted Instruction · Educational technology · Electronic learning · Statistical analysis · Student achievement · Teaching method · Expert finding and Q&A systems · Intelligent Tutoring Systems and Adaptive Learning · Online Learning and Analytics

Tools

Open DOI