Students view AI videos as appropriate for simple, visualizable, and supplemental learning contexts but inappropriate for complex, subjective, or interactive ones¶

Claim · All claims
Evidence · 1 study · qualitative r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Thematic analysis of 163 open-ended responses identified appropriate avenues (simple topics, visualizations, summaries, supplemental material) and inappropriate avenues (complex topics, subjective topics, replacing lectures, interactive learning). → Esse Ciego 2026

Evidence¶

Esse Ciego 2026¶

Esse Ciego, Shubbhi Taneja, Wilson Wong, and Amanpreet Kapoor. (2026). Student Perceptions and Preferences Regarding AI-Generated Instructional Videos in Computing Education. https://arxiv.org/abs/2607.28203 doi:10.48550/arxiv.2607.28203

q2 · i? · qualitative · r2

Inductive thematic analysis of open-ended survey responses from 163 students. Appropriate themes included simple topics (n=44, 27%), topics with visualizations (n=24, 15%), topic summaries (n=20, 12%), and supplemental material (n=20, 12%); inappropriate themes included complex topics (n=50, 31%) and replacing lectures (n=26, 16%).

"We coded 163 non-blank responses to an open-ended question "Describe one educational scenario where AI-generated videos would work well and one where they would not. Explain your reasoning for each. "into 14 unique codes abstracted into 5 themes."

Discussion¶

Related Claims¶

• Guo et al. found engagement depends on video production style: short videos, talking heads on slides, Khan-style drawing more engaging than slides or coding, classroom recording weak online — the general claim this finding bears on, with the studies behind it
• Only half of students could clearly tell the AI-generated videos were AI-generated, the item with the highest variance — related
• Students' primary concerns about AI videos are inaccurate information, reduced interaction with educators, and diminished educational value — related
• Most U.S. colleges had adopted written generative AI policies by late 2024, with 69% defining appropriate versus inappropriate AI use in coursework — related
• Most surveyed college leaders report written GenAI responsible-use policies, but decision-making is largely left to educators — related
• Letting learners pace instruction helped on complex material in one experiment, while learner control in general showed almost no effect across 18 studies — related
• Qualitative themes: efficiency and peer influence drive AI use, while confidence and rule ambiguity shape students' perceptions of dependence — related
• AI-mediated SSH learning externalizes interpretive cognition as students delegate thematic identification, contextual explanation, and comparative analysis to AI systems — related
• Qualitative reflections identify developmental inappropriateness (45%), vagueness and repetition (32%), and implementation feasibility (23%) as the main AI limitations — related
• LLMs perform worse on complex, open-ended, or subjective assignments than on short, clearly formulated answers — related
• MI Write's automated feedback is most useful to students when it supplements, not replaces, teacher feedback — a narrower finding that bears on this claim
• Task complexity moderates whether users experience the constructive information search process — related